Skip to content

feat(server): /mcp narrows a client's modes from T3-Mode-Limit - #16653

Merged
juliusmarminge merged 1 commit into
mainfrom
t3code/peer/mode-limit-header
Oct 27, 2026
Merged

juliusmarminge merged 1 commit into
mainfrom
t3code/peer/mode-limit-header

Conversation

@juliusmarminge

@juliusmarminge juliusmarminge commented Oct 7, 2026 •

Copy link
Copy Markdown
Member

Part of cross-environment orchestration: an environment linked to another signs in to its /mcp as an outside agent. Every linked call has to carry the calling agent's own limits, so a plan-mode agent on the laptop can't start full-access work on the box just because the link was approved at auto.

What changes

  • /mcp reads an optional T3-Mode-Limit: <runtime>/<interaction> header, from OAuth (mcp-client) callers only.
  • The header can only narrow the client's ceiling. A broader value changes nothing, and a malformed one gets a 400 (invalid_mode_limit).
  • clientModeCeiling replaces clientRuntimeModeCeiling, adding an interaction ceiling. loadCaller and the orchestrator MCP service use it, so the access declarations and the in-lock DispatchModeLimit check enforce the narrowed modes.
  • Provider sessions ignore the header; their limits come from their thread.

Verification

  • New test in McpHttpServer.test.ts. A real /mcp, with an auto OAuth client, serves a probe tool that starts a thread through McpToolAccess.startsThreads. Results:
    • no header → auto/default;
    • approval-required/plan → narrowed, and escalation requests above it are refused;
    • a broader header → unchanged;
    • malformed → 400.
  • Mutation-checked both ways: dropping the narrowing fails the test, and so does letting a broader header widen the ceiling.
  • MCP and auth suites: 37 files, 364 tests, all passing.

Opus 5.5 via Claude Code.

🤖 Generated with Claude Code


Devin Review

@github-actions github-actions Bot added vouch:trusted PR author is trusted by repo permissions or the VOUCHED list. size:L 100-499 changed lines (additions + deletions). labels Oct 7, 2026
@juliusmarminge juliusmarminge changed the title t3code/peer/mode limit header feat(server): /mcp narrows a client's modes from T3-Mode-Limit Oct 7, 2026
@juliusmarminge juliusmarminge added the macroscope-review Opt PRs made by unvouched contributors in for Macroscope review. Vouched contributors auto-reviews label Oct 7, 2026
@github-actions

github-actions Bot commented Oct 7, 2026 •

Copy link
Copy Markdown
Contributor

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider Metric Main baseline This PR Impact PR ceiling
Codex Total thread wire 5.0 KiB 5.0 KiB −6 B (−0.1%) 6.8 KiB ✅
Codex Thread snapshot wire 3.8 KiB 3.8 KiB 0 B (0.0%) 4.9 KiB ✅
Codex Live turn WebSocket wire 1.3 KiB 1.2 KiB −6 B (−0.5%) 2.0 KiB ✅
Codex Live turn WebSocket decoded 21.6 KiB 21.6 KiB 0 B (0.0%) 24.4 KiB ✅
Codex Live turn messages 2 2 0 (0.0%) 8 ✅
Claude Total thread wire 5.1 KiB 5.0 KiB −6 B (−0.1%) 6.8 KiB ✅
Claude Thread snapshot wire 3.8 KiB 3.8 KiB 0 B (0.0%) 4.9 KiB ✅
Claude Live turn WebSocket wire 1.3 KiB 1.3 KiB −6 B (−0.5%) 2.0 KiB ✅
Claude Live turn WebSocket decoded 21.9 KiB 21.9 KiB 0 B (0.0%) 24.4 KiB ✅
Claude Live turn messages 2 2 0 (0.0%) 8 ✅

Baseline: 5808469 · PR result: 898a813 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

  • Codex decoded thread snapshot: 108.5 KiB
  • Claude decoded thread snapshot: 108.8 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

@juliusmarminge

Copy link
Copy Markdown
Member Author

End-to-end run, two real servers

Two t3 serve processes from the top of this stack, each with its own data directory, on one machine. A laptop (port 3971) and a box (port 3972). Their projects are clones of one bare origin, so they share a repository. An outside agent (OAuth with a pairing code) drives the laptop's /mcp, and real Claude Sonnet 5.5 turns run on both sides. The last part repeats the run over Tailscale HTTPS (a *.ts.net HTTPS address).

Through the link, the box refused launches above the caller's modes. A plain call at full-access got "Child runtime mode full-access is broader than parent mode auto." (the link's access was auto). The same launch at auto with T3-Mode-Limit: approval-required/plan got "Child runtime mode auto is broader than parent mode approval-required.", so the header narrowed the ceiling below the link's.

CI also showed an unused MODE_LIMIT_HEADER export (Knip). The constant now lives in McpInvocationContext, where #16684 already uses it.

Opus 5.5 via Claude Code.

@juliusmarminge
juliusmarminge marked this pull request as ready for review October 7, 2026 18:09
@coderabbitai

coderabbitai Bot commented Oct 7, 2026 •

Copy link
Copy Markdown
Contributor

Review in Change Stack →

Note

Reviews paused

It looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the reviews.auto_review.auto_pause_after_reviewed_commits setting.

Use the following commands to manage reviews:

  • @coderabbitai resume to resume automatic reviews.
  • @coderabbitai review to trigger a single review.

Use the checkboxes below for quick actions:

  • ▶️ Resume reviews
  • 🔍 Trigger review

No actionable comments were generated in the recent review.

ℹ️ Recent review info
⚙️ Run configuration
  • Configuration used: Path: .coderabbit.config.ts
  • Review profile: CHILL
  • Plan: Team
  • Run ID: d94a2616-bd0d-4db2-a460-cf472ccc8b20
📥 Commits

Reviewing files that changed from the base of the PR and between 8eadb66 and 0b07846.

📒 Files selected for processing (1)
  • apps/server/src/mcp/OrchestratorMcpService.ts

Included review availability: This review used your included allowance. Your plan provides up to 10 included reviews per hour; 7 remain after this review.

Walkthrough

The MCP HTTP server accepts a T3-Mode-Limit header for authenticated clients. It validates the requested modes and applies them as limits to the client’s runtime and interaction modes. MCP callers without a thread use the resulting ceilings.

Changes: OAuth Client Mode Limits

Layer / File(s) Summary
Request mode-limit contract and validation
apps/server/src/mcp/McpInvocationContext.ts, apps/server/src/mcp/McpHttpServer.ts
The invocation context defines the mode-limit header and optional narrowed modes. The HTTP server validates the header and returns HTTP 400 for malformed or unsupported values.
Effective mode ceilings and callers
apps/server/src/mcp/McpInvocationContext.ts, apps/server/src/mcp/OrchestratorMcpService.ts, apps/server/src/mcp/threadAccess.ts, apps/server/src/mcp/McpHttpServer.test.ts
clientModeCeiling independently applies runtime and interaction limits. Callers without a thread use those ceilings. The integration test checks narrowing, non-widening, escalation-denied responses, and invalid headers.

Priority: Normal

Estimated code review effort: 3 (Moderate) | ~20 minutes

Change: Feature

Sequence Diagram(s)

sequenceDiagram
  participant OAuthClient
  participant McpHttpServer
  participant OrchestratorMcpService
  OAuthClient->>McpHttpServer: Send request with T3-Mode-Limit
  McpHttpServer->>OrchestratorMcpService: Pass invocation context to handler
  OrchestratorMcpService->>OrchestratorMcpService: Calculate runtime and interaction ceilings
Loading

Suggested reviewers: maria-rcks

Merge Risk: Minimal · previous review 0b078

The change narrows OAuth MCP callers’ mode limits, and both limits reach dispatch enforcement. No concrete merge-blocking risk is apparent.

🚥 Pre-merge checks | ✅ 3 | ❌ 1

❌ Failed checks (1 warning)

Check name Status Explanation Resolution
Description check ⚠️ Warning The description explains the problem, implementation, affected callers, and verification results. It does not include the required Scope and approval section or an issue, approval reference, or vali… Add a ## Scope and approval section. Link the triaged issue or maintainer approval with the approval comment, or explain why this focused change qualifies for an exemption.
✅ Passed checks (3 passed)
Check name Status Explanation
Title check ✅ Passed The title clearly describes the main change: /mcp narrows client modes using T3-Mode-Limit.
Linked Issues check ✅ Passed Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check ✅ Passed Check skipped because no linked issues were found for this pull request.
Full details: Description check

Explanation

The description explains the problem, implementation, affected callers, and verification results. It does not include the required Scope and approval section or an issue, approval reference, or valid exemption rationale.

✨ Finishing Touches
📝 Generate docstrings
  • Commit to this branch
  • Create a new PR
🧪 Generate unit tests (beta)
  • Commit to this branch
  • Create a new PR
  • Autopilot · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Comment @coderabbitai help to get the list of available commands.

@macroscopeapp

macroscopeapp Bot commented Oct 7, 2026 •

Copy link
Copy Markdown
Contributor

Approvability

Verdict: Needs Human Approval

Macroscope's review found this PR requires human approval — The PR adds an optional T3-Mode-Limit header that changes MCP authorization ceilings and constrains downstream thread modes. It also introduces a new static-analysis suppression in the test, making this more than a routine additive or mechanical change.

No code changes detected at 898a813. Prior analysis still applies.

You can add or adjust custom eligibility rules. Learn more.

An OAuth client on /mcp is capped by the access it was approved with.
When another environment calls on behalf of one of its agents, that agent
may run under narrower modes, and the target has no way to know.

/mcp now reads `T3-Mode-Limit: <runtimeMode>/<interactionMode>` from OAuth
clients and caps the request at the narrower of the two. The header can
only narrow: a broader one changes nothing, and a malformed one is
refused with 400. The client's interaction mode can now be capped too;
before, it was always default.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
@juliusmarminge
juliusmarminge force-pushed the t3code/peer/mode-limit-header branch from ed9f466 to 898a813 Compare October 27, 2026 20:49
@juliusmarminge
juliusmarminge merged commit 1b1ad04 into main Oct 27, 2026
31 checks passed
@juliusmarminge
juliusmarminge deleted the t3code/peer/mode-limit-header branch October 27, 2026 21:04
github-actions Bot added a commit to omarcresp/t3code-flake that referenced this pull request Oct 28, 2026
## What's Changed
* fix(mobile): center Android sheets and fit book-mode sheets in the left pane by @PixPMusic in pingdotgg/t3code#18492
* fix(mobile): keep the collapsed Android rail square by @PixPMusic in pingdotgg/t3code#26039
* fix(web): closing the last Browser tab hides the right panel again by @juliusmarminge in pingdotgg/t3code#26058
* refactor(web): pull requests Cmd+F focuses the topbar search directly by @juliusmarminge in pingdotgg/t3code#26045
* refactor(web): install draft model reconciliation at app start by @juliusmarminge in pingdotgg/t3code#26046
* perf(mobile): project favicons mount their image in one commit by @juliusmarminge in pingdotgg/t3code#26047
* refactor(web): the upload queue alone records finished file uploads on drafts by @juliusmarminge in pingdotgg/t3code#26050
* feat(mobile): Usage uses the width on iPad and unfolded phones by @juliusmarminge in pingdotgg/t3code#26061
* fix(gitcafe): PR panel follows renumbered pull requests by @juliusmarminge in pingdotgg/t3code#26079
* chore(infra): bump Alchemy to 2.0.0-beta.84 by @juliusmarminge in pingdotgg/t3code#26057
* refactor(mobile): stop writing the resolved project filter back from an effect by @juliusmarminge in pingdotgg/t3code#26049
* fix(mobile): a pairing link fills the open Add environment form instead of stacking another by @juliusmarminge in pingdotgg/t3code#26052
* fix(mobile): snooze menu times advance with the clock by @juliusmarminge in pingdotgg/t3code#26043
* fix(clients): prompt links find projects cloned from git.cafe by @juliusmarminge in pingdotgg/t3code#26042
* fix(server): retry Codex capacity failures with empty input by @Yash-Singh1 in pingdotgg/t3code#25611
* fix(server): recover handoffs when target context is exhausted by @Yash-Singh1 in pingdotgg/t3code#25610
* fix(web): chips drop their pill and stay readable on every theme by @flamboh in pingdotgg/t3code#25066
* fix(mobile): switching pull request tabs no longer jumps the page by @flamboh in pingdotgg/t3code#26064
* feat(mobile): report crashes to Axiom with symbolicated stacks by @gabrielelpidio in pingdotgg/t3code#19093
* fix(web): chip palette test follows the single design by @juliusmarminge in pingdotgg/t3code#26131
* feat(server): /mcp narrows a client's modes from T3-Mode-Limit by @juliusmarminge in pingdotgg/t3code#16653
* feat(server): t3_thread_launch takes clientRequestId by @juliusmarminge in pingdotgg/t3code#16654
* feat(server): link to another environment by @juliusmarminge in pingdotgg/t3code#16655
* feat(server): agents use linked environments through any T3 tool by @juliusmarminge in pingdotgg/t3code#16684
* feat(server): work a linked environment starts stays within its link by @juliusmarminge in pingdotgg/t3code#16695
* feat(web,mobile): linked environments in Connections by @juliusmarminge in pingdotgg/t3code#16698
* feat(server): delegate_task to a linked environment by @juliusmarminge in pingdotgg/t3code#16719
* feat(server,web,mobile): an agent sets up a link to another environment by @juliusmarminge in pingdotgg/t3code#17129
* feat(server,web,mobile): @-mention a machine, and linked environments become a Labs beta by @juliusmarminge in pingdotgg/t3code#17130
* fix(mobile): dragging a microphone or environment route now saves its new place by @juliusmarminge in pingdotgg/t3code#26096
* feat(web): channel artwork choice on Latest, plus nightly or dev art on any channel by @maria-rcks in pingdotgg/t3code#26138
* feat(mobile): browse ZIP attachments with the shared archive reader by @Bil0000 in pingdotgg/t3code#19463
* fix(server): Claude --settings launch args merge with T3's settings instead of being dropped by @t3dotgg in pingdotgg/t3code#22864
* feat(server,web): see and stop the dev servers a thread started by @t3dotgg in pingdotgg/t3code#22542
* fix(web): composer no longer jumps when a second line is added by @maria-rcks in pingdotgg/t3code#26155
* feat(web,mobile): queued messages show who sent them by @juliusmarminge in pingdotgg/t3code#26159
* fix(preview): agent and browser panel show the same page after a late desktop attach by @t3dotgg in pingdotgg/t3code#22877
* fix(web): pairing, connect, error, and Clerk sign-in surfaces match the app's dialog style by @maria-rcks in pingdotgg/t3code#25281
* fix(server): parallel worktree launches no longer fail on .git/config lock by @juliusmarminge in pingdotgg/t3code#25700
* fix(task-cache): the hook test waits for every flusher it starts by @juliusmarminge in pingdotgg/t3code#26148
* chore(lint): every React effect states its reason at the call site by @juliusmarminge in pingdotgg/t3code#26161
* test(server): the link-card retry test names a new MCP credential by @juliusmarminge in pingdotgg/t3code#26143
* perf(web): stop measuring table cells while replies stream by @t3dotgg in pingdotgg/t3code#26152
* feat(desktop): add a remote server by clicking its t3code://pair link by @t3dotgg in pingdotgg/t3code#24111
* fix(server): agent questions stay answerable after a server restart by @SunkenInTime in pingdotgg/t3code#19229
* chore(lint): the effect lint error walks the alternatives before the suppression by @juliusmarminge in pingdotgg/t3code#26181
* fix(web): settings avatar no longer hugs the modal corner by @t3dotgg in pingdotgg/t3code#26180
* fix(mobile): shared file names can no longer write outside the Android cache by @juliusmarminge in pingdotgg/t3code#26175
* fix(server): repository lookups no longer send GitHub tokens to arbitrary hosts by @juliusmarminge in pingdotgg/t3code#26176
* fix(github): long period runs no longer freeze reference autolinking by @juliusmarminge in pingdotgg/t3code#26177
* fix(mobile): Android diff ellipsis no longer measures text quadratically by @juliusmarminge in pingdotgg/t3code#26174
* fix(server): Codex works again after its app-server is killed by @t3dotgg in pingdotgg/t3code#26188


**Full Changelog**: pingdotgg/t3code@v0.1.4-nightly.20261027.3783...v0.1.4-nightly.20261027.3796

Upstream release: https://github.com/pingdotgg/t3code/releases/tag/v0.1.4-nightly.20261027.3796
Yarden-zamir added a commit to Yarden-zamir/t3code that referenced this pull request Oct 28, 2026
Upstream merged the adopted linked environments stack (pingdotgg#16653 to pingdotgg#16719).
Conflicts take upstream's stack code, then apply the fork hooks again: arrows give the
links (MachinePeers), the CLI refuses manual links, and Settings shows arrow links.
immakermatty added a commit to Lazurio/t3code that referenced this pull request Oct 29, 2026
…r decided a request

Root decision 0210, evening addendum, point 5 (plan DEV-6676, row 27): what
Buddy writes or decides in a thread carries the visible mark „Buddy za <user>“
("Buddy for <user>"). Delegace's wake in Buddy's own thread is not shown at all,
as on Telegram (decided 2026-10-22), so the person sees only Buddy's reply.

Who a session acts as is a claim signed into its token beside the runtime-mode
ceiling (`act`), from a closed set of kinds: buddy with the user's name, or
delegace; budicek is no alias. The agent commands issue their sessions with it
when T3CODE_ACTING_SENDER (with T3CODE_ACTING_FOR and
T3CODE_ACTING_FOR_ACCUSATIVE) names one, and refuse any other value before they
read anything. The server reads it from the session alone: an MCP client's
invocation carries it, the client surface runs a dispatch as its session, and a
decorator of the orchestrator records the message or the request's resolution
in a file per thread in the T3 home, only when its own dispatch committed that
step. A replayed command id, an overlapping dispatch of the same id or a retry
takes no one else's step, and a step's first record stands. The web reads
/api/lazurio/acting-senders. The thread's rows leave out Delegace's wake and
hold an agent's message from outside the thread until the answer covers it,
five seconds at most; a server without the route, or an error, shows the thread
at once as upstream does. Buddy's message and a request Buddy decided show
„Buddy za …“. The person's own steps, T3's messages, automations and other
threads' agents get no mark and never wait. Tests: rejections before
successes, the real CLI against /mcp and the client surface through to the
route the web reads, replays on upstream's own orchestrator, the claim in the
token, and the web's label, parse and loader.

On v0.1.3 three small conflicts are resolved by intent. Upstream computes the
folded agent message's sender label once (pingdotgg#24104) and also reads it out in
the accessible heading (pingdotgg#23780); the overlay wraps that senderLabel in
LazurioActingSenderLabel on the visible fold, which is the one place the
mark goes. McpOAuth's authenticate is upstream's Effect.fn now (pingdotgg#25099), and
the client gets the one actingSender spread line there. EnvironmentAuth keeps
row 20's MCP_CLIENT_SUBJECT export. Upstream's guest attribution (pingdotgg#24104)
is a message-context record of its own and does not mark who acted for a
person, so the overlay's session claim stays. The orchestrator decorator's
dispatch is an Effect.fn (effectFnOpportunity), the two effects that ask
the server's record carry justified react-effect-budget suppressions, the
agent CLI test keeps row 21's queue and fork case beside this row's cases,
and long comments are cut to the four-line limit.

Buddy's day pane counts the messages the timeline shows: the server's day list
(buddySurfaceDays.ts) leaves out Delegace's wake and an agent's message the
record has not answered yet, by the predicate in actingSenderVisibility.ts,
which the server and the web hold as one identical file that the release
contract test keeps equal. A day that opens with a wake jumps to the person's
first message, and a day of wakes alone is not listed.

On v0.1.4: the acting sender sits beside upstream's peer link (pingdotgg#16695) in the session, its signed claims, the MCP client and its caller, next to upstream's narrowed modes and web origin (pingdotgg#16653, pingdotgg#26642); the test takes mcpClientScopes from contracts, where upstream moved it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
immakermatty added a commit to Lazurio/t3code that referenced this pull request Oct 30, 2026
…r decided a request

Root decision 0210, evening addendum, point 5 (plan DEV-6676, row 27): what
Buddy writes or decides in a thread carries the visible mark „Buddy za <user>“
("Buddy for <user>"). Delegace's wake in Buddy's own thread is not shown at all,
as on Telegram (decided 2026-10-22), so the person sees only Buddy's reply.

Who a session acts as is a claim signed into its token beside the runtime-mode
ceiling (`act`), from a closed set of kinds: buddy with the user's name, or
delegace; budicek is no alias. The agent commands issue their sessions with it
when T3CODE_ACTING_SENDER (with T3CODE_ACTING_FOR and
T3CODE_ACTING_FOR_ACCUSATIVE) names one, and refuse any other value before they
read anything. The server reads it from the session alone: an MCP client's
invocation carries it, the client surface runs a dispatch as its session, and a
decorator of the orchestrator records the message or the request's resolution
in a file per thread in the T3 home, only when its own dispatch committed that
step. A replayed command id, an overlapping dispatch of the same id or a retry
takes no one else's step, and a step's first record stands. The web reads
/api/lazurio/acting-senders. The thread's rows leave out Delegace's wake and
hold an agent's message from outside the thread until the answer covers it,
five seconds at most; a server without the route, or an error, shows the thread
at once as upstream does. Buddy's message and a request Buddy decided show
„Buddy za …“. The person's own steps, T3's messages, automations and other
threads' agents get no mark and never wait. Tests: rejections before
successes, the real CLI against /mcp and the client surface through to the
route the web reads, replays on upstream's own orchestrator, the claim in the
token, and the web's label, parse and loader.

On v0.1.3 three small conflicts are resolved by intent. Upstream computes the
folded agent message's sender label once (pingdotgg#24104) and also reads it out in
the accessible heading (pingdotgg#23780); the overlay wraps that senderLabel in
LazurioActingSenderLabel on the visible fold, which is the one place the
mark goes. McpOAuth's authenticate is upstream's Effect.fn now (pingdotgg#25099), and
the client gets the one actingSender spread line there. EnvironmentAuth keeps
row 20's MCP_CLIENT_SUBJECT export. Upstream's guest attribution (pingdotgg#24104)
is a message-context record of its own and does not mark who acted for a
person, so the overlay's session claim stays. The orchestrator decorator's
dispatch is an Effect.fn (effectFnOpportunity), the two effects that ask
the server's record carry justified react-effect-budget suppressions, the
agent CLI test keeps row 21's queue and fork case beside this row's cases,
and long comments are cut to the four-line limit.

Buddy's day pane counts the messages the timeline shows: the server's day list
(buddySurfaceDays.ts) leaves out Delegace's wake and an agent's message the
record has not answered yet, by the predicate in actingSenderVisibility.ts,
which the server and the web hold as one identical file that the release
contract test keeps equal. A day that opens with a wake jumps to the person's
first message, and a day of wakes alone is not listed.

On v0.1.4: the acting sender sits beside upstream's peer link (pingdotgg#16695) in the session, its signed claims, the MCP client and its caller, next to upstream's narrowed modes and web origin (pingdotgg#16653, pingdotgg#26642); the test takes mcpClientScopes from contracts, where upstream moved it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
immakermatty added a commit to Lazurio/t3code that referenced this pull request Oct 30, 2026
…r decided a request

Root decision 0210, evening addendum, point 5 (plan DEV-6676, row 27): what
Buddy writes or decides in a thread carries the visible mark „Buddy za <user>“
("Buddy for <user>"). Delegace's wake in Buddy's own thread is not shown at all,
as on Telegram (decided 2026-10-22), so the person sees only Buddy's reply.

Who a session acts as is a claim signed into its token beside the runtime-mode
ceiling (`act`), from a closed set of kinds: buddy with the user's name, or
delegace; budicek is no alias. The agent commands issue their sessions with it
when T3CODE_ACTING_SENDER (with T3CODE_ACTING_FOR and
T3CODE_ACTING_FOR_ACCUSATIVE) names one, and refuse any other value before they
read anything. The server reads it from the session alone: an MCP client's
invocation carries it, the client surface runs a dispatch as its session, and a
decorator of the orchestrator records the message or the request's resolution
in a file per thread in the T3 home, only when its own dispatch committed that
step. A replayed command id, an overlapping dispatch of the same id or a retry
takes no one else's step, and a step's first record stands. The web reads
/api/lazurio/acting-senders. The thread's rows leave out Delegace's wake and
hold an agent's message from outside the thread until the answer covers it,
five seconds at most; a server without the route, or an error, shows the thread
at once as upstream does. Buddy's message and a request Buddy decided show
„Buddy za …“. The person's own steps, T3's messages, automations and other
threads' agents get no mark and never wait. Tests: rejections before
successes, the real CLI against /mcp and the client surface through to the
route the web reads, replays on upstream's own orchestrator, the claim in the
token, and the web's label, parse and loader.

On v0.1.3 three small conflicts are resolved by intent. Upstream computes the
folded agent message's sender label once (pingdotgg#24104) and also reads it out in
the accessible heading (pingdotgg#23780); the overlay wraps that senderLabel in
LazurioActingSenderLabel on the visible fold, which is the one place the
mark goes. McpOAuth's authenticate is upstream's Effect.fn now (pingdotgg#25099), and
the client gets the one actingSender spread line there. EnvironmentAuth keeps
row 20's MCP_CLIENT_SUBJECT export. Upstream's guest attribution (pingdotgg#24104)
is a message-context record of its own and does not mark who acted for a
person, so the overlay's session claim stays. The orchestrator decorator's
dispatch is an Effect.fn (effectFnOpportunity), the two effects that ask
the server's record carry justified react-effect-budget suppressions, the
agent CLI test keeps row 21's queue and fork case beside this row's cases,
and long comments are cut to the four-line limit.

Buddy's day pane counts the messages the timeline shows: the server's day list
(buddySurfaceDays.ts) leaves out Delegace's wake and an agent's message the
record has not answered yet, by the predicate in actingSenderVisibility.ts,
which the server and the web hold as one identical file that the release
contract test keeps equal. A day that opens with a wake jumps to the person's
first message, and a day of wakes alone is not listed.

On v0.1.4: the acting sender sits beside upstream's peer link (pingdotgg#16695) in the session, its signed claims, the MCP client and its caller, next to upstream's narrowed modes and web origin (pingdotgg#16653, pingdotgg#26642); the test takes mcpClientScopes from contracts, where upstream moved it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
immakermatty added a commit to Lazurio/t3code that referenced this pull request Oct 30, 2026
…r decided a request

Root decision 0210, evening addendum, point 5 (plan DEV-6676, row 27): what
Buddy writes or decides in a thread carries the visible mark „Buddy za <user>“
("Buddy for <user>"). Delegace's wake in Buddy's own thread is not shown at all,
as on Telegram (decided 2026-10-22), so the person sees only Buddy's reply.

Who a session acts as is a claim signed into its token beside the runtime-mode
ceiling (`act`), from a closed set of kinds: buddy with the user's name, or
delegace; budicek is no alias. The agent commands issue their sessions with it
when T3CODE_ACTING_SENDER (with T3CODE_ACTING_FOR and
T3CODE_ACTING_FOR_ACCUSATIVE) names one, and refuse any other value before they
read anything. The server reads it from the session alone: an MCP client's
invocation carries it, the client surface runs a dispatch as its session, and a
decorator of the orchestrator records the message or the request's resolution
in a file per thread in the T3 home, only when its own dispatch committed that
step. A replayed command id, an overlapping dispatch of the same id or a retry
takes no one else's step, and a step's first record stands. The web reads
/api/lazurio/acting-senders. The thread's rows leave out Delegace's wake and
hold an agent's message from outside the thread until the answer covers it,
five seconds at most; a server without the route, or an error, shows the thread
at once as upstream does. Buddy's message and a request Buddy decided show
„Buddy za …“. The person's own steps, T3's messages, automations and other
threads' agents get no mark and never wait. Tests: rejections before
successes, the real CLI against /mcp and the client surface through to the
route the web reads, replays on upstream's own orchestrator, the claim in the
token, and the web's label, parse and loader.

On v0.1.3 three small conflicts are resolved by intent. Upstream computes the
folded agent message's sender label once (pingdotgg#24104) and also reads it out in
the accessible heading (pingdotgg#23780); the overlay wraps that senderLabel in
LazurioActingSenderLabel on the visible fold, which is the one place the
mark goes. McpOAuth's authenticate is upstream's Effect.fn now (pingdotgg#25099), and
the client gets the one actingSender spread line there. EnvironmentAuth keeps
row 20's MCP_CLIENT_SUBJECT export. Upstream's guest attribution (pingdotgg#24104)
is a message-context record of its own and does not mark who acted for a
person, so the overlay's session claim stays. The orchestrator decorator's
dispatch is an Effect.fn (effectFnOpportunity), the two effects that ask
the server's record carry justified react-effect-budget suppressions (now
t3code/no-unreasoned-state-sync), the
agent CLI test keeps row 21's queue and fork case beside this row's cases,
and long comments are cut to the four-line limit.

Buddy's day pane counts the messages the timeline shows: the server's day list
(buddySurfaceDays.ts) leaves out Delegace's wake and an agent's message the
record has not answered yet, by the predicate in actingSenderVisibility.ts,
which the server and the web hold as one identical file that the release
contract test keeps equal. A day that opens with a wake jumps to the person's
first message, and a day of wakes alone is not listed.

On v0.1.4: the acting sender sits beside upstream's peer link (pingdotgg#16695) in the session, its signed claims, the MCP client and its caller, next to upstream's narrowed modes and web origin (pingdotgg#16653, pingdotgg#26642); the test takes mcpClientScopes from contracts, where upstream moved it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
immakermatty added a commit to Lazurio/t3code that referenced this pull request Oct 30, 2026
…r decided a request

Root decision 0210, evening addendum, point 5 (plan DEV-6676, row 27): what
Buddy writes or decides in a thread carries the visible mark „Buddy za <user>“
("Buddy for <user>"). Delegace's wake in Buddy's own thread is not shown at all,
as on Telegram (decided 2026-10-22), so the person sees only Buddy's reply.

Who a session acts as is a claim signed into its token beside the runtime-mode
ceiling (`act`), from a closed set of kinds: buddy with the user's name, or
delegace; budicek is no alias. The agent commands issue their sessions with it
when T3CODE_ACTING_SENDER (with T3CODE_ACTING_FOR and
T3CODE_ACTING_FOR_ACCUSATIVE) names one, and refuse any other value before they
read anything. The server reads it from the session alone: an MCP client's
invocation carries it, the client surface runs a dispatch as its session, and a
decorator of the orchestrator records the message or the request's resolution
in a file per thread in the T3 home, only when its own dispatch committed that
step. A replayed command id, an overlapping dispatch of the same id or a retry
takes no one else's step, and a step's first record stands. The web reads
/api/lazurio/acting-senders. The thread's rows leave out Delegace's wake and
hold an agent's message from outside the thread until the answer covers it,
five seconds at most; a server without the route, or an error, shows the thread
at once as upstream does. Buddy's message and a request Buddy decided show
„Buddy za …“. The person's own steps, T3's messages, automations and other
threads' agents get no mark and never wait. Tests: rejections before
successes, the real CLI against /mcp and the client surface through to the
route the web reads, replays on upstream's own orchestrator, the claim in the
token, and the web's label, parse and loader.

On v0.1.3 three small conflicts are resolved by intent. Upstream computes the
folded agent message's sender label once (pingdotgg#24104) and also reads it out in
the accessible heading (pingdotgg#23780); the overlay wraps that senderLabel in
LazurioActingSenderLabel on the visible fold, which is the one place the
mark goes. McpOAuth's authenticate is upstream's Effect.fn now (pingdotgg#25099), and
the client gets the one actingSender spread line there. EnvironmentAuth keeps
row 20's MCP_CLIENT_SUBJECT export. Upstream's guest attribution (pingdotgg#24104)
is a message-context record of its own and does not mark who acted for a
person, so the overlay's session claim stays. The orchestrator decorator's
dispatch is an Effect.fn (effectFnOpportunity), the two effects that ask
the server's record carry justified react-effect-budget suppressions (now
t3code/no-unreasoned-state-sync), the
agent CLI test keeps row 21's queue and fork case beside this row's cases,
and long comments are cut to the four-line limit.

Buddy's day pane counts the messages the timeline shows: the server's day list
(buddySurfaceDays.ts) leaves out Delegace's wake and an agent's message the
record has not answered yet, by the predicate in actingSenderVisibility.ts,
which the server and the web hold as one identical file that the release
contract test keeps equal. A day that opens with a wake jumps to the person's
first message, and a day of wakes alone is not listed.

On v0.1.4: the acting sender sits beside upstream's peer link (pingdotgg#16695) in the session, its signed claims, the MCP client and its caller, next to upstream's narrowed modes and web origin (pingdotgg#16653, pingdotgg#26642); the test takes mcpClientScopes from contracts, where upstream moved it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
immakermatty added a commit to Lazurio/t3code that referenced this pull request Oct 30, 2026
…r decided a request

Root decision 0210, evening addendum, point 5 (plan DEV-6676, row 27): what
Buddy writes or decides in a thread carries the visible mark „Buddy za <user>“
("Buddy for <user>"). Delegace's wake in Buddy's own thread is not shown at all,
as on Telegram (decided 2026-10-22), so the person sees only Buddy's reply.

Who a session acts as is a claim signed into its token beside the runtime-mode
ceiling (`act`), from a closed set of kinds: buddy with the user's name, or
delegace; budicek is no alias. The agent commands issue their sessions with it
when T3CODE_ACTING_SENDER (with T3CODE_ACTING_FOR and
T3CODE_ACTING_FOR_ACCUSATIVE) names one, and refuse any other value before they
read anything. The server reads it from the session alone: an MCP client's
invocation carries it, the client surface runs a dispatch as its session, and a
decorator of the orchestrator records the message or the request's resolution
in a file per thread in the T3 home, only when its own dispatch committed that
step. A replayed command id, an overlapping dispatch of the same id or a retry
takes no one else's step, and a step's first record stands. The web reads
/api/lazurio/acting-senders. The thread's rows leave out Delegace's wake and
hold an agent's message from outside the thread until the answer covers it,
five seconds at most; a server without the route, or an error, shows the thread
at once as upstream does. Buddy's message and a request Buddy decided show
„Buddy za …“. The person's own steps, T3's messages, automations and other
threads' agents get no mark and never wait. Tests: rejections before
successes, the real CLI against /mcp and the client surface through to the
route the web reads, replays on upstream's own orchestrator, the claim in the
token, and the web's label, parse and loader.

On v0.1.3 three small conflicts are resolved by intent. Upstream computes the
folded agent message's sender label once (pingdotgg#24104) and also reads it out in
the accessible heading (pingdotgg#23780); the overlay wraps that senderLabel in
LazurioActingSenderLabel on the visible fold, which is the one place the
mark goes. McpOAuth's authenticate is upstream's Effect.fn now (pingdotgg#25099), and
the client gets the one actingSender spread line there. EnvironmentAuth keeps
row 20's MCP_CLIENT_SUBJECT export. Upstream's guest attribution (pingdotgg#24104)
is a message-context record of its own and does not mark who acted for a
person, so the overlay's session claim stays. The orchestrator decorator's
dispatch is an Effect.fn (effectFnOpportunity), the two effects that ask
the server's record carry justified react-effect-budget suppressions (now
t3code/no-unreasoned-state-sync), the
agent CLI test keeps row 21's queue and fork case beside this row's cases,
and long comments are cut to the four-line limit.

Buddy's day pane counts the messages the timeline shows: the server's day list
(buddySurfaceDays.ts) leaves out Delegace's wake and an agent's message the
record has not answered yet, by the predicate in actingSenderVisibility.ts,
which the server and the web hold as one identical file that the release
contract test keeps equal. A day that opens with a wake jumps to the person's
first message, and a day of wakes alone is not listed.

On v0.1.4: the acting sender sits beside upstream's peer link (pingdotgg#16695) in the session, its signed claims, the MCP client and its caller, next to upstream's narrowed modes and web origin (pingdotgg#16653, pingdotgg#26642); the test takes mcpClientScopes from contracts, where upstream moved it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
immakermatty added a commit to Lazurio/t3code that referenced this pull request Oct 30, 2026
…r decided a request

Root decision 0210, evening addendum, point 5 (plan DEV-6676, row 27): what
Buddy writes or decides in a thread carries the visible mark „Buddy za <user>“
("Buddy for <user>"). Delegace's wake in Buddy's own thread is not shown at all,
as on Telegram (decided 2026-10-22), so the person sees only Buddy's reply.

Who a session acts as is a claim signed into its token beside the runtime-mode
ceiling (`act`), from a closed set of kinds: buddy with the user's name, or
delegace; budicek is no alias. The agent commands issue their sessions with it
when T3CODE_ACTING_SENDER (with T3CODE_ACTING_FOR and
T3CODE_ACTING_FOR_ACCUSATIVE) names one, and refuse any other value before they
read anything. The server reads it from the session alone: an MCP client's
invocation carries it, the client surface runs a dispatch as its session, and a
decorator of the orchestrator records the message or the request's resolution
in a file per thread in the T3 home, only when its own dispatch committed that
step. A replayed command id, an overlapping dispatch of the same id or a retry
takes no one else's step, and a step's first record stands. The web reads
/api/lazurio/acting-senders. The thread's rows leave out Delegace's wake and
hold an agent's message from outside the thread until the answer covers it,
five seconds at most; a server without the route, or an error, shows the thread
at once as upstream does. Buddy's message and a request Buddy decided show
„Buddy za …“. The person's own steps, T3's messages, automations and other
threads' agents get no mark and never wait. Tests: rejections before
successes, the real CLI against /mcp and the client surface through to the
route the web reads, replays on upstream's own orchestrator, the claim in the
token, and the web's label, parse and loader.

On v0.1.3 three small conflicts are resolved by intent. Upstream computes the
folded agent message's sender label once (pingdotgg#24104) and also reads it out in
the accessible heading (pingdotgg#23780); the overlay wraps that senderLabel in
LazurioActingSenderLabel on the visible fold, which is the one place the
mark goes. McpOAuth's authenticate is upstream's Effect.fn now (pingdotgg#25099), and
the client gets the one actingSender spread line there. EnvironmentAuth keeps
row 20's MCP_CLIENT_SUBJECT export. Upstream's guest attribution (pingdotgg#24104)
is a message-context record of its own and does not mark who acted for a
person, so the overlay's session claim stays. The orchestrator decorator's
dispatch is an Effect.fn (effectFnOpportunity), the two effects that ask
the server's record carry justified react-effect-budget suppressions
(now t3code/no-unreasoned-state-sync), the agent CLI test keeps row 21's
queue and fork case beside this row's cases, and long comments are cut to
the four-line limit.

Buddy's day pane counts the messages the timeline shows: the server's day list
(buddySurfaceDays.ts) leaves out Delegace's wake and an agent's message the
record has not answered yet, by the predicate in actingSenderVisibility.ts,
which the server and the web hold as one identical file that the release
contract test keeps equal. A day that opens with a wake jumps to the person's
first message, and a day of wakes alone is not listed.

On v0.1.4: the acting sender sits beside upstream's peer link (pingdotgg#16695) in the session, its signed claims, the MCP client and its caller, next to upstream's narrowed modes and web origin (pingdotgg#16653, pingdotgg#26642); the test takes mcpClientScopes from contracts, where upstream moved it.

On v0.1.5 upstream's pre-commit hook runs knip (pingdotgg#26167), so
ACTING_SENDER_NAME_MAX_LENGTH, ClientSurfaceActingSender,
actingSenders.ts's make and ACTING_SENDERS_ROUTE, which nothing imports,
are module-private. actingSenderVisibility.ts must stay the same file on
the server and in the web, and each side uses part of it, so its
functions keep their exports and are @public to knip.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
immakermatty added a commit to Lazurio/t3code that referenced this pull request Oct 31, 2026
…r decided a request

Root decision 0210, evening addendum, point 5 (plan DEV-6676, row 27): what
Buddy writes or decides in a thread carries the visible mark „Buddy za <user>“
("Buddy for <user>"). Delegace's wake in Buddy's own thread is not shown at all,
as on Telegram (decided 2026-10-22), so the person sees only Buddy's reply.

Who a session acts as is a claim signed into its token beside the runtime-mode
ceiling (`act`), from a closed set of kinds: buddy with the user's name, or
delegace; budicek is no alias. The agent commands issue their sessions with it
when T3CODE_ACTING_SENDER (with T3CODE_ACTING_FOR and
T3CODE_ACTING_FOR_ACCUSATIVE) names one, and refuse any other value before they
read anything. The server reads it from the session alone: an MCP client's
invocation carries it, the client surface runs a dispatch as its session, and a
decorator of the orchestrator records the message or the request's resolution
in a file per thread in the T3 home, only when its own dispatch committed that
step. A replayed command id, an overlapping dispatch of the same id or a retry
takes no one else's step, and a step's first record stands. The web reads
/api/lazurio/acting-senders. The thread's rows leave out Delegace's wake and
hold an agent's message from outside the thread until the answer covers it,
five seconds at most; a server without the route, or an error, shows the thread
at once as upstream does. Buddy's message and a request Buddy decided show
„Buddy za …“. The person's own steps, T3's messages, automations and other
threads' agents get no mark and never wait. Tests: rejections before
successes, the real CLI against /mcp and the client surface through to the
route the web reads, replays on upstream's own orchestrator, the claim in the
token, and the web's label, parse and loader.

On v0.1.3 three small conflicts are resolved by intent. Upstream computes the
folded agent message's sender label once (pingdotgg#24104) and also reads it out in
the accessible heading (pingdotgg#23780); the overlay wraps that senderLabel in
LazurioActingSenderLabel on the visible fold, which is the one place the
mark goes. McpOAuth's authenticate is upstream's Effect.fn now (pingdotgg#25099), and
the client gets the one actingSender spread line there. EnvironmentAuth keeps
row 20's MCP_CLIENT_SUBJECT export. Upstream's guest attribution (pingdotgg#24104)
is a message-context record of its own and does not mark who acted for a
person, so the overlay's session claim stays. The orchestrator decorator's
dispatch is an Effect.fn (effectFnOpportunity), the two effects that ask
the server's record carry justified react-effect-budget suppressions
(now t3code/no-unreasoned-state-sync), the agent CLI test keeps row 21's
queue and fork case beside this row's cases, and long comments are cut to
the four-line limit.

Buddy's day pane counts the messages the timeline shows: the server's day list
(buddySurfaceDays.ts) leaves out Delegace's wake and an agent's message the
record has not answered yet, by the predicate in actingSenderVisibility.ts,
which the server and the web hold as one identical file that the release
contract test keeps equal. A day that opens with a wake jumps to the person's
first message, and a day of wakes alone is not listed.

On v0.1.4: the acting sender sits beside upstream's peer link (pingdotgg#16695) in the session, its signed claims, the MCP client and its caller, next to upstream's narrowed modes and web origin (pingdotgg#16653, pingdotgg#26642); the test takes mcpClientScopes from contracts, where upstream moved it.

On v0.1.5 upstream's pre-commit hook runs knip (pingdotgg#26167), so
ACTING_SENDER_NAME_MAX_LENGTH, ClientSurfaceActingSender,
actingSenders.ts's make and ACTING_SENDERS_ROUTE, which nothing imports,
are module-private. actingSenderVisibility.ts must stay the same file on
the server and in the web, and each side uses part of it, so its
functions keep their exports and are @public to knip.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
immakermatty added a commit to Lazurio/t3code that referenced this pull request Nov 4, 2026
…r decided a request

Root decision 0210, evening addendum, point 5 (plan DEV-6676, row 27): what
Buddy writes or decides in a thread carries the visible mark „Buddy za <user>“
("Buddy for <user>"). Delegace's wake in Buddy's own thread is not shown at all,
as on Telegram (decided 2026-10-22), so the person sees only Buddy's reply.

Who a session acts as is a claim signed into its token beside the runtime-mode
ceiling (`act`), from a closed set of kinds: buddy with the user's name, or
delegace; budicek is no alias. The agent commands issue their sessions with it
when T3CODE_ACTING_SENDER (with T3CODE_ACTING_FOR and
T3CODE_ACTING_FOR_ACCUSATIVE) names one, and refuse any other value before they
read anything. The server reads it from the session alone: an MCP client's
invocation carries it, the client surface runs a dispatch as its session, and a
decorator of the orchestrator records the message or the request's resolution
in a file per thread in the T3 home, only when its own dispatch committed that
step. A replayed command id, an overlapping dispatch of the same id or a retry
takes no one else's step, and a step's first record stands. The web reads
/api/lazurio/acting-senders. The thread's rows leave out Delegace's wake and
hold an agent's message from outside the thread until the answer covers it,
five seconds at most; a server without the route, or an error, shows the thread
at once as upstream does. Buddy's message and a request Buddy decided show
„Buddy za …“. The person's own steps, T3's messages, automations and other
threads' agents get no mark and never wait. Tests: rejections before
successes, the real CLI against /mcp and the client surface through to the
route the web reads, replays on upstream's own orchestrator, the claim in the
token, and the web's label, parse and loader.

On v0.1.3 three small conflicts are resolved by intent. Upstream computes the
folded agent message's sender label once (pingdotgg#24104) and also reads it out in
the accessible heading (pingdotgg#23780); the overlay wraps that senderLabel in
LazurioActingSenderLabel on the visible fold, which is the one place the
mark goes. McpOAuth's authenticate is upstream's Effect.fn now (pingdotgg#25099), and
the client gets the one actingSender spread line there. EnvironmentAuth keeps
row 20's MCP_CLIENT_SUBJECT export. Upstream's guest attribution (pingdotgg#24104)
is a message-context record of its own and does not mark who acted for a
person, so the overlay's session claim stays. The orchestrator decorator's
dispatch is an Effect.fn (effectFnOpportunity), the two effects that ask
the server's record carry justified react-effect-budget suppressions
(now t3code/no-unreasoned-state-sync), the agent CLI test keeps row 21's
queue and fork case beside this row's cases, and long comments are cut to
the four-line limit.

Buddy's day pane counts the messages the timeline shows: the server's day list
(buddySurfaceDays.ts) leaves out Delegace's wake and an agent's message the
record has not answered yet, by the predicate in actingSenderVisibility.ts,
which the server and the web hold as one identical file that the release
contract test keeps equal. A day that opens with a wake jumps to the person's
first message, and a day of wakes alone is not listed.

On v0.1.4: the acting sender sits beside upstream's peer link (pingdotgg#16695) in the session, its signed claims, the MCP client and its caller, next to upstream's narrowed modes and web origin (pingdotgg#16653, pingdotgg#26642); the test takes mcpClientScopes from contracts, where upstream moved it.

On v0.1.5 upstream's pre-commit hook runs knip (pingdotgg#26167), so
ACTING_SENDER_NAME_MAX_LENGTH, ClientSurfaceActingSender,
actingSenders.ts's make and ACTING_SENDERS_ROUTE, which nothing imports,
are module-private. actingSenderVisibility.ts must stay the same file on
the server and in the web, and each side uses part of it, so its
functions keep their exports and are @public to knip.

On v0.1.7: upstream renamed the MCP caller's `webOrigin` to
`hostedAppUrl` and sets it in the OAuth authenticator for every client
that is not a peer link. The acting sender rides beside it unchanged. The
auth tests provide `ManagedPolicy.layerTest()`, which `EnvironmentAuth`
now needs, and `NodeCrypto.layer`, which Effect 4.1.2 no longer brings
with `NodeServices.layer`.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
immakermatty added a commit to Lazurio/t3code that referenced this pull request Nov 4, 2026
…r decided a request

Root decision 0210, evening addendum, point 5 (plan DEV-6676, row 27): what
Buddy writes or decides in a thread carries the visible mark „Buddy za <user>“
("Buddy for <user>"). Delegace's wake in Buddy's own thread is not shown at all,
as on Telegram (decided 2026-10-22), so the person sees only Buddy's reply.

Who a session acts as is a claim signed into its token beside the runtime-mode
ceiling (`act`), from a closed set of kinds: buddy with the user's name, or
delegace; budicek is no alias. The agent commands issue their sessions with it
when T3CODE_ACTING_SENDER (with T3CODE_ACTING_FOR and
T3CODE_ACTING_FOR_ACCUSATIVE) names one, and refuse any other value before they
read anything. The server reads it from the session alone: an MCP client's
invocation carries it, the client surface runs a dispatch as its session, and a
decorator of the orchestrator records the message or the request's resolution
in a file per thread in the T3 home, only when its own dispatch committed that
step. A replayed command id, an overlapping dispatch of the same id or a retry
takes no one else's step, and a step's first record stands. The web reads
/api/lazurio/acting-senders. The thread's rows leave out Delegace's wake and
hold an agent's message from outside the thread until the answer covers it,
five seconds at most; a server without the route, or an error, shows the thread
at once as upstream does. Buddy's message and a request Buddy decided show
„Buddy za …“. The person's own steps, T3's messages, automations and other
threads' agents get no mark and never wait. Tests: rejections before
successes, the real CLI against /mcp and the client surface through to the
route the web reads, replays on upstream's own orchestrator, the claim in the
token, and the web's label, parse and loader.

On v0.1.3 three small conflicts are resolved by intent. Upstream computes the
folded agent message's sender label once (pingdotgg#24104) and also reads it out in
the accessible heading (pingdotgg#23780); the overlay wraps that senderLabel in
LazurioActingSenderLabel on the visible fold, which is the one place the
mark goes. McpOAuth's authenticate is upstream's Effect.fn now (pingdotgg#25099), and
the client gets the one actingSender spread line there. EnvironmentAuth keeps
row 20's MCP_CLIENT_SUBJECT export. Upstream's guest attribution (pingdotgg#24104)
is a message-context record of its own and does not mark who acted for a
person, so the overlay's session claim stays. The orchestrator decorator's
dispatch is an Effect.fn (effectFnOpportunity), the two effects that ask
the server's record carry justified react-effect-budget suppressions
(now t3code/no-unreasoned-state-sync), the agent CLI test keeps row 21's
queue and fork case beside this row's cases, and long comments are cut to
the four-line limit.

Buddy's day pane counts the messages the timeline shows: the server's day list
(buddySurfaceDays.ts) leaves out Delegace's wake and an agent's message the
record has not answered yet, by the predicate in actingSenderVisibility.ts,
which the server and the web hold as one identical file that the release
contract test keeps equal. A day that opens with a wake jumps to the person's
first message, and a day of wakes alone is not listed.

On v0.1.4: the acting sender sits beside upstream's peer link (pingdotgg#16695) in the session, its signed claims, the MCP client and its caller, next to upstream's narrowed modes and web origin (pingdotgg#16653, pingdotgg#26642); the test takes mcpClientScopes from contracts, where upstream moved it.

On v0.1.5 upstream's pre-commit hook runs knip (pingdotgg#26167), so
ACTING_SENDER_NAME_MAX_LENGTH, ClientSurfaceActingSender,
actingSenders.ts's make and ACTING_SENDERS_ROUTE, which nothing imports,
are module-private. actingSenderVisibility.ts must stay the same file on
the server and in the web, and each side uses part of it, so its
functions keep their exports and are @public to knip.

On v0.1.7: upstream renamed the MCP caller's `webOrigin` to
`hostedAppUrl` and sets it in the OAuth authenticator for every client
that is not a peer link. The acting sender rides beside it unchanged. The
auth tests provide `ManagedPolicy.layerTest()`, which `EnvironmentAuth`
now needs, and `NodeCrypto.layer`, which Effect 4.1.2 no longer brings
with `NodeServices.layer`.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
immakermatty added a commit to Lazurio/t3code that referenced this pull request Nov 5, 2026
…r decided a request

Root decision 0210, evening addendum, point 5 (plan DEV-6676, row 27): what
Buddy writes or decides in a thread carries the visible mark „Buddy za <user>“
("Buddy for <user>"). Delegace's wake in Buddy's own thread is not shown at all,
as on Telegram (decided 2026-10-22), so the person sees only Buddy's reply.

Who a session acts as is a claim signed into its token beside the runtime-mode
ceiling (`act`), from a closed set of kinds: buddy with the user's name, or
delegace; budicek is no alias. The agent commands issue their sessions with it
when T3CODE_ACTING_SENDER (with T3CODE_ACTING_FOR and
T3CODE_ACTING_FOR_ACCUSATIVE) names one, and refuse any other value before they
read anything. The server reads it from the session alone: an MCP client's
invocation carries it, the client surface runs a dispatch as its session, and a
decorator of the orchestrator records the message or the request's resolution
in a file per thread in the T3 home, only when its own dispatch committed that
step. A replayed command id, an overlapping dispatch of the same id or a retry
takes no one else's step, and a step's first record stands. The web reads
/api/lazurio/acting-senders. The thread's rows leave out Delegace's wake and
hold an agent's message from outside the thread until the answer covers it,
five seconds at most; a server without the route, or an error, shows the thread
at once as upstream does. Buddy's message and a request Buddy decided show
„Buddy za …“. The person's own steps, T3's messages, automations and other
threads' agents get no mark and never wait. Tests: rejections before
successes, the real CLI against /mcp and the client surface through to the
route the web reads, replays on upstream's own orchestrator, the claim in the
token, and the web's label, parse and loader.

On v0.1.3 three small conflicts are resolved by intent. Upstream computes the
folded agent message's sender label once (pingdotgg#24104) and also reads it out in
the accessible heading (pingdotgg#23780); the overlay wraps that senderLabel in
LazurioActingSenderLabel on the visible fold, which is the one place the
mark goes. McpOAuth's authenticate is upstream's Effect.fn now (pingdotgg#25099), and
the client gets the one actingSender spread line there. EnvironmentAuth keeps
row 20's MCP_CLIENT_SUBJECT export. Upstream's guest attribution (pingdotgg#24104)
is a message-context record of its own and does not mark who acted for a
person, so the overlay's session claim stays. The orchestrator decorator's
dispatch is an Effect.fn (effectFnOpportunity), the two effects that ask
the server's record carry justified react-effect-budget suppressions
(now t3code/no-unreasoned-state-sync), the agent CLI test keeps row 21's
queue and fork case beside this row's cases, and long comments are cut to
the four-line limit.

Buddy's day pane counts the messages the timeline shows: the server's day list
(buddySurfaceDays.ts) leaves out Delegace's wake and an agent's message the
record has not answered yet, by the predicate in actingSenderVisibility.ts,
which the server and the web hold as one identical file that the release
contract test keeps equal. A day that opens with a wake jumps to the person's
first message, and a day of wakes alone is not listed.

On v0.1.4: the acting sender sits beside upstream's peer link (pingdotgg#16695) in the session, its signed claims, the MCP client and its caller, next to upstream's narrowed modes and web origin (pingdotgg#16653, pingdotgg#26642); the test takes mcpClientScopes from contracts, where upstream moved it.

On v0.1.5 upstream's pre-commit hook runs knip (pingdotgg#26167), so
ACTING_SENDER_NAME_MAX_LENGTH, ClientSurfaceActingSender,
actingSenders.ts's make and ACTING_SENDERS_ROUTE, which nothing imports,
are module-private. actingSenderVisibility.ts must stay the same file on
the server and in the web, and each side uses part of it, so its
functions keep their exports and are @public to knip.

On v0.1.7: upstream renamed the MCP caller's `webOrigin` to
`hostedAppUrl` and sets it in the OAuth authenticator for every client
that is not a peer link. The acting sender rides beside it unchanged. The
auth tests provide `ManagedPolicy.layerTest()`, which `EnvironmentAuth`
now needs, and `NodeCrypto.layer`, which Effect 4.1.2 no longer brings
with `NodeServices.layer`.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
immakermatty added a commit to Lazurio/t3code that referenced this pull request Nov 5, 2026
…r decided a request

Root decision 0210, evening addendum, point 5 (plan DEV-6676, row 27): what
Buddy writes or decides in a thread carries the visible mark „Buddy za <user>“
("Buddy for <user>"). Delegace's wake in Buddy's own thread is not shown at all,
as on Telegram (decided 2026-10-22), so the person sees only Buddy's reply.

Who a session acts as is a claim signed into its token beside the runtime-mode
ceiling (`act`), from a closed set of kinds: buddy with the user's name, or
delegace; budicek is no alias. The agent commands issue their sessions with it
when T3CODE_ACTING_SENDER (with T3CODE_ACTING_FOR and
T3CODE_ACTING_FOR_ACCUSATIVE) names one, and refuse any other value before they
read anything. The server reads it from the session alone: an MCP client's
invocation carries it, the client surface runs a dispatch as its session, and a
decorator of the orchestrator records the message or the request's resolution
in a file per thread in the T3 home, only when its own dispatch committed that
step. A replayed command id, an overlapping dispatch of the same id or a retry
takes no one else's step, and a step's first record stands. The web reads
/api/lazurio/acting-senders. The thread's rows leave out Delegace's wake and
hold an agent's message from outside the thread until the answer covers it,
five seconds at most; a server without the route, or an error, shows the thread
at once as upstream does. Buddy's message and a request Buddy decided show
„Buddy za …“. The person's own steps, T3's messages, automations and other
threads' agents get no mark and never wait. Tests: rejections before
successes, the real CLI against /mcp and the client surface through to the
route the web reads, replays on upstream's own orchestrator, the claim in the
token, and the web's label, parse and loader.

On v0.1.3 three small conflicts are resolved by intent. Upstream computes the
folded agent message's sender label once (pingdotgg#24104) and also reads it out in
the accessible heading (pingdotgg#23780); the overlay wraps that senderLabel in
LazurioActingSenderLabel on the visible fold, which is the one place the
mark goes. McpOAuth's authenticate is upstream's Effect.fn now (pingdotgg#25099), and
the client gets the one actingSender spread line there. EnvironmentAuth keeps
row 20's MCP_CLIENT_SUBJECT export. Upstream's guest attribution (pingdotgg#24104)
is a message-context record of its own and does not mark who acted for a
person, so the overlay's session claim stays. The orchestrator decorator's
dispatch is an Effect.fn (effectFnOpportunity), the two effects that ask
the server's record carry justified react-effect-budget suppressions
(now t3code/no-unreasoned-state-sync), the agent CLI test keeps row 21's
queue and fork case beside this row's cases, and long comments are cut to
the four-line limit.

Buddy's day pane counts the messages the timeline shows: the server's day list
(buddySurfaceDays.ts) leaves out Delegace's wake and an agent's message the
record has not answered yet, by the predicate in actingSenderVisibility.ts,
which the server and the web hold as one identical file that the release
contract test keeps equal. A day that opens with a wake jumps to the person's
first message, and a day of wakes alone is not listed.

On v0.1.4: the acting sender sits beside upstream's peer link (pingdotgg#16695) in the session, its signed claims, the MCP client and its caller, next to upstream's narrowed modes and web origin (pingdotgg#16653, pingdotgg#26642); the test takes mcpClientScopes from contracts, where upstream moved it.

On v0.1.5 upstream's pre-commit hook runs knip (pingdotgg#26167), so
ACTING_SENDER_NAME_MAX_LENGTH, ClientSurfaceActingSender,
actingSenders.ts's make and ACTING_SENDERS_ROUTE, which nothing imports,
are module-private. actingSenderVisibility.ts must stay the same file on
the server and in the web, and each side uses part of it, so its
functions keep their exports and are @public to knip.

On v0.1.7: upstream renamed the MCP caller's `webOrigin` to
`hostedAppUrl` and sets it in the OAuth authenticator for every client
that is not a peer link. The acting sender rides beside it unchanged. The
auth tests provide `ManagedPolicy.layerTest()`, which `EnvironmentAuth`
now needs, and `NodeCrypto.layer`, which Effect 4.1.2 no longer brings
with `NodeServices.layer`.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
immakermatty added a commit to Lazurio/t3code that referenced this pull request Nov 5, 2026
…r decided a request

Root decision 0210, evening addendum, point 5 (plan DEV-6676, row 27): what
Buddy writes or decides in a thread carries the visible mark „Buddy za <user>“
("Buddy for <user>"). Delegace's wake in Buddy's own thread is not shown at all,
as on Telegram (decided 2026-10-22), so the person sees only Buddy's reply.

Who a session acts as is a claim signed into its token beside the runtime-mode
ceiling (`act`), from a closed set of kinds: buddy with the user's name, or
delegace; budicek is no alias. The agent commands issue their sessions with it
when T3CODE_ACTING_SENDER (with T3CODE_ACTING_FOR and
T3CODE_ACTING_FOR_ACCUSATIVE) names one, and refuse any other value before they
read anything. The server reads it from the session alone: an MCP client's
invocation carries it, the client surface runs a dispatch as its session, and a
decorator of the orchestrator records the message or the request's resolution
in a file per thread in the T3 home, only when its own dispatch committed that
step. A replayed command id, an overlapping dispatch of the same id or a retry
takes no one else's step, and a step's first record stands. The web reads
/api/lazurio/acting-senders. The thread's rows leave out Delegace's wake and
hold an agent's message from outside the thread until the answer covers it,
five seconds at most; a server without the route, or an error, shows the thread
at once as upstream does. Buddy's message and a request Buddy decided show
„Buddy za …“. The person's own steps, T3's messages, automations and other
threads' agents get no mark and never wait. Tests: rejections before
successes, the real CLI against /mcp and the client surface through to the
route the web reads, replays on upstream's own orchestrator, the claim in the
token, and the web's label, parse and loader.

On v0.1.3 three small conflicts are resolved by intent. Upstream computes the
folded agent message's sender label once (pingdotgg#24104) and also reads it out in
the accessible heading (pingdotgg#23780); the overlay wraps that senderLabel in
LazurioActingSenderLabel on the visible fold, which is the one place the
mark goes. McpOAuth's authenticate is upstream's Effect.fn now (pingdotgg#25099), and
the client gets the one actingSender spread line there. EnvironmentAuth keeps
row 20's MCP_CLIENT_SUBJECT export. Upstream's guest attribution (pingdotgg#24104)
is a message-context record of its own and does not mark who acted for a
person, so the overlay's session claim stays. The orchestrator decorator's
dispatch is an Effect.fn (effectFnOpportunity), the two effects that ask
the server's record carry justified react-effect-budget suppressions
(now t3code/no-unreasoned-state-sync), the agent CLI test keeps row 21's
queue and fork case beside this row's cases, and long comments are cut to
the four-line limit.

Buddy's day pane counts the messages the timeline shows: the server's day list
(buddySurfaceDays.ts) leaves out Delegace's wake and an agent's message the
record has not answered yet, by the predicate in actingSenderVisibility.ts,
which the server and the web hold as one identical file that the release
contract test keeps equal. A day that opens with a wake jumps to the person's
first message, and a day of wakes alone is not listed.

On v0.1.4: the acting sender sits beside upstream's peer link (pingdotgg#16695) in the session, its signed claims, the MCP client and its caller, next to upstream's narrowed modes and web origin (pingdotgg#16653, pingdotgg#26642); the test takes mcpClientScopes from contracts, where upstream moved it.

On v0.1.5 upstream's pre-commit hook runs knip (pingdotgg#26167), so
ACTING_SENDER_NAME_MAX_LENGTH, ClientSurfaceActingSender,
actingSenders.ts's make and ACTING_SENDERS_ROUTE, which nothing imports,
are module-private. actingSenderVisibility.ts must stay the same file on
the server and in the web, and each side uses part of it, so its
functions keep their exports and are @public to knip.

On v0.1.7: upstream renamed the MCP caller's `webOrigin` to
`hostedAppUrl` and sets it in the OAuth authenticator for every client
that is not a peer link. The acting sender rides beside it unchanged. The
auth tests provide `ManagedPolicy.layerTest()`, which `EnvironmentAuth`
now needs, and `NodeCrypto.layer`, which Effect 4.1.2 no longer brings
with `NodeServices.layer`.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
immakermatty added a commit to Lazurio/t3code that referenced this pull request Nov 5, 2026
…r decided a request

Root decision 0210, evening addendum, point 5 (plan DEV-6676, row 27): what
Buddy writes or decides in a thread carries the visible mark „Buddy za <user>“
("Buddy for <user>"). Delegace's wake in Buddy's own thread is not shown at all,
as on Telegram (decided 2026-10-22), so the person sees only Buddy's reply.

Who a session acts as is a claim signed into its token beside the runtime-mode
ceiling (`act`), from a closed set of kinds: buddy with the user's name, or
delegace; budicek is no alias. The agent commands issue their sessions with it
when T3CODE_ACTING_SENDER (with T3CODE_ACTING_FOR and
T3CODE_ACTING_FOR_ACCUSATIVE) names one, and refuse any other value before they
read anything. The server reads it from the session alone: an MCP client's
invocation carries it, the client surface runs a dispatch as its session, and a
decorator of the orchestrator records the message or the request's resolution
in a file per thread in the T3 home, only when its own dispatch committed that
step. A replayed command id, an overlapping dispatch of the same id or a retry
takes no one else's step, and a step's first record stands. The web reads
/api/lazurio/acting-senders. The thread's rows leave out Delegace's wake and
hold an agent's message from outside the thread until the answer covers it,
five seconds at most; a server without the route, or an error, shows the thread
at once as upstream does. Buddy's message and a request Buddy decided show
„Buddy za …“. The person's own steps, T3's messages, automations and other
threads' agents get no mark and never wait. Tests: rejections before
successes, the real CLI against /mcp and the client surface through to the
route the web reads, replays on upstream's own orchestrator, the claim in the
token, and the web's label, parse and loader.

On v0.1.3 three small conflicts are resolved by intent. Upstream computes the
folded agent message's sender label once (pingdotgg#24104) and also reads it out in
the accessible heading (pingdotgg#23780); the overlay wraps that senderLabel in
LazurioActingSenderLabel on the visible fold, which is the one place the
mark goes. McpOAuth's authenticate is upstream's Effect.fn now (pingdotgg#25099), and
the client gets the one actingSender spread line there. EnvironmentAuth keeps
row 20's MCP_CLIENT_SUBJECT export. Upstream's guest attribution (pingdotgg#24104)
is a message-context record of its own and does not mark who acted for a
person, so the overlay's session claim stays. The orchestrator decorator's
dispatch is an Effect.fn (effectFnOpportunity), the two effects that ask
the server's record carry justified react-effect-budget suppressions
(now t3code/no-unreasoned-state-sync), the agent CLI test keeps row 21's
queue and fork case beside this row's cases, and long comments are cut to
the four-line limit.

Buddy's day pane counts the messages the timeline shows: the server's day list
(buddySurfaceDays.ts) leaves out Delegace's wake and an agent's message the
record has not answered yet, by the predicate in actingSenderVisibility.ts,
which the server and the web hold as one identical file that the release
contract test keeps equal. A day that opens with a wake jumps to the person's
first message, and a day of wakes alone is not listed.

On v0.1.4: the acting sender sits beside upstream's peer link (pingdotgg#16695) in the session, its signed claims, the MCP client and its caller, next to upstream's narrowed modes and web origin (pingdotgg#16653, pingdotgg#26642); the test takes mcpClientScopes from contracts, where upstream moved it.

On v0.1.5 upstream's pre-commit hook runs knip (pingdotgg#26167), so
ACTING_SENDER_NAME_MAX_LENGTH, ClientSurfaceActingSender,
actingSenders.ts's make and ACTING_SENDERS_ROUTE, which nothing imports,
are module-private. actingSenderVisibility.ts must stay the same file on
the server and in the web, and each side uses part of it, so its
functions keep their exports and are @public to knip.

On v0.1.7: upstream renamed the MCP caller's `webOrigin` to
`hostedAppUrl` and sets it in the OAuth authenticator for every client
that is not a peer link. The acting sender rides beside it unchanged. The
auth tests provide `ManagedPolicy.layerTest()`, which `EnvironmentAuth`
now needs, and `NodeCrypto.layer`, which Effect 4.1.2 no longer brings
with `NodeServices.layer`.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
immakermatty added a commit to Lazurio/t3code that referenced this pull request Nov 5, 2026
…r decided a request

Root decision 0210, evening addendum, point 5 (plan DEV-6676, row 27): what
Buddy writes or decides in a thread carries the visible mark „Buddy za <user>“
("Buddy for <user>"). Delegace's wake in Buddy's own thread is not shown at all,
as on Telegram (decided 2026-10-22), so the person sees only Buddy's reply.

Who a session acts as is a claim signed into its token beside the runtime-mode
ceiling (`act`), from a closed set of kinds: buddy with the user's name, or
delegace; budicek is no alias. The agent commands issue their sessions with it
when T3CODE_ACTING_SENDER (with T3CODE_ACTING_FOR and
T3CODE_ACTING_FOR_ACCUSATIVE) names one, and refuse any other value before they
read anything. The server reads it from the session alone: an MCP client's
invocation carries it, the client surface runs a dispatch as its session, and a
decorator of the orchestrator records the message or the request's resolution
in a file per thread in the T3 home, only when its own dispatch committed that
step. A replayed command id, an overlapping dispatch of the same id or a retry
takes no one else's step, and a step's first record stands. The web reads
/api/lazurio/acting-senders. The thread's rows leave out Delegace's wake and
hold an agent's message from outside the thread until the answer covers it,
five seconds at most; a server without the route, or an error, shows the thread
at once as upstream does. Buddy's message and a request Buddy decided show
„Buddy za …“. The person's own steps, T3's messages, automations and other
threads' agents get no mark and never wait. Tests: rejections before
successes, the real CLI against /mcp and the client surface through to the
route the web reads, replays on upstream's own orchestrator, the claim in the
token, and the web's label, parse and loader.

On v0.1.3 three small conflicts are resolved by intent. Upstream computes the
folded agent message's sender label once (pingdotgg#24104) and also reads it out in
the accessible heading (pingdotgg#23780); the overlay wraps that senderLabel in
LazurioActingSenderLabel on the visible fold, which is the one place the
mark goes. McpOAuth's authenticate is upstream's Effect.fn now (pingdotgg#25099), and
the client gets the one actingSender spread line there. EnvironmentAuth keeps
row 20's MCP_CLIENT_SUBJECT export. Upstream's guest attribution (pingdotgg#24104)
is a message-context record of its own and does not mark who acted for a
person, so the overlay's session claim stays. The orchestrator decorator's
dispatch is an Effect.fn (effectFnOpportunity), the two effects that ask
the server's record carry justified react-effect-budget suppressions
(now t3code/no-unreasoned-state-sync), the agent CLI test keeps row 21's
queue and fork case beside this row's cases, and long comments are cut to
the four-line limit.

Buddy's day pane counts the messages the timeline shows: the server's day list
(buddySurfaceDays.ts) leaves out Delegace's wake and an agent's message the
record has not answered yet, by the predicate in actingSenderVisibility.ts,
which the server and the web hold as one identical file that the release
contract test keeps equal. A day that opens with a wake jumps to the person's
first message, and a day of wakes alone is not listed.

On v0.1.4: the acting sender sits beside upstream's peer link (pingdotgg#16695) in the session, its signed claims, the MCP client and its caller, next to upstream's narrowed modes and web origin (pingdotgg#16653, pingdotgg#26642); the test takes mcpClientScopes from contracts, where upstream moved it.

On v0.1.5 upstream's pre-commit hook runs knip (pingdotgg#26167), so
ACTING_SENDER_NAME_MAX_LENGTH, ClientSurfaceActingSender,
actingSenders.ts's make and ACTING_SENDERS_ROUTE, which nothing imports,
are module-private. actingSenderVisibility.ts must stay the same file on
the server and in the web, and each side uses part of it, so its
functions keep their exports and are @public to knip.

On v0.1.7: upstream renamed the MCP caller's `webOrigin` to
`hostedAppUrl` and sets it in the OAuth authenticator for every client
that is not a peer link. The acting sender rides beside it unchanged. The
auth tests provide `ManagedPolicy.layerTest()`, which `EnvironmentAuth`
now needs, and `NodeCrypto.layer`, which Effect 4.1.2 no longer brings
with `NodeServices.layer`.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

macroscope-review Opt PRs made by unvouched contributors in for Macroscope review. Vouched contributors auto-reviews size:L 100-499 changed lines (additions + deletions). vouch:trusted PR author is trusted by repo permissions or the VOUCHED list.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants