Skip to content

docs(operations): the wake delivers the caller's reason, not its payload - #137

Merged
Fl0p merged 1 commit into
mainfrom
flo-998-wake-origin-doc
Oct 5, 2026
Merged

Fl0p merged 1 commit into
mainfrom
flo-998-wake-origin-doc

Conversation

@Fl0p

@Fl0p Fl0p commented Oct 5, 2026

Copy link
Copy Markdown
Contributor

The pager section claimed the free-text reason is bucketed to an enum before it reaches the woken agent. Reading the wake path shows it is not: wakeup's reason lands on the run's context as wakeReason and is delivered three ways - the PAPERCLIP_WAKE_REASON environment variable, the reason field of PAPERCLIP_WAKE_PAYLOAD_JSON, and a rendered - reason: ... line in the wake summary. The enum bucketing belongs to the diagnostics/wakes response projection, which collapses an unrecognised reason to other - that is the artifact a drill reads, not what the agent gets.

Corroborated on the 2026-10-05 pair: the run the green half's wake was absorbed into carries, on its stored context, the pager's full sentence naming the recovery and the green run URL.

Two gaps are real and are different from the one the doc named:

  • The wake's source is not copied into the agent's payload at all, so no delivered field says a wake came from automation.
  • reason does not survive coalescing. The payload a run is handed is built once, at dispatch; a wake arriving afterwards is absorbed without reaching it. That is why the assignee read a continuation label while the absorbed wake row carried the pager's sentence.

So reason is a second channel, not a substitute - delivered on the ordinary production shape where a standing alert has no live run, swallowed in exactly the case these drills keep landing in. The description stays the only channel that survives both.

Docs only; no behaviour change. Verification is the Docs workflow building VitePress on this PR.

The pager section claimed the free-text `reason` is bucketed to an enum before
it reaches the woken agent. It is not: it is carried verbatim into the run's
context and arrives as `PAPERCLIP_WAKE_REASON`, as `reason` in
`PAPERCLIP_WAKE_PAYLOAD_JSON`, and as a rendered line in the wake summary. The
bucketing is done by the `diagnostics/wakes` response projection, which is
what a drill reads, not the path to the agent.

Two gaps are real, and they are different ones. The wake's `source` is never
copied into the agent's payload, so no delivered field says a wake came from
automation. And a wake that arrives after a run has been dispatched is absorbed
into it without reaching it, which is what the 2026-10-05 pair observed: the
assignee read a continuation label while the absorbed wake row carried the
pager's sentence. That makes `reason` a second channel rather than a
substitute, and leaves the description as the only one that survives both.

Co-Authored-By: Daedalus <daedalus@agents.flopbut.local>
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
@coderabbitai

coderabbitai Bot commented Oct 5, 2026

Copy link
Copy Markdown

Warning

Review limit reached

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

Next included review available in 54 minutes.

Check out review usage here.

View limit details

Limit details: You’ve used the included review currently available.

Learn how review limits work.

Review configuration:

⚙️ Run configuration
  • Configuration used: defaults
  • Review profile: CHILL
  • Plan: Advanced
  • Run ID: eb289c89-92c6-4ab2-8c10-e75f269a6fee
📥 Commits

Reviewing files that changed from the base of the PR and between fee77a2 and 97d4188.

📒 Files selected for processing (1)
  • docs/operations/health-probe.md
  • Autopilot · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

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

@Fl0p
Fl0p merged commit 9cc7c43 into main Oct 5, 2026
8 checks passed
@Fl0p
Fl0p deleted the flo-998-wake-origin-doc branch October 5, 2026 09:51
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant