docs(changelog): tag-based runner placement (Jun 06)#80
Merged
Conversation
Agents now route to runners by capability: SKILL.md `tags:` become an agent's required_tags, and a runner claims an agent only when those tags are a subset of its advertised labels. Empty tags run anywhere (back-compat). Notes the silent-wait when no runner matches and the UZ-REQ-001 bounds. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
|
Preview deployment for your docs. Learn more about Mintlify Previews.
💡 Tip: Enable Workflows to automatically generate PRs for you. |
Address greptile review on PR #80: - Split the conflated tag bounds in the API reference bullet — per-tag length (1–64 chars) and per-agent count (≤32) are distinct triggers for UZ-REQ-001; "either bound exceeded" now reads unambiguously. - Add a <Note> on the silent-wait behavior. A waiting agent surfaces no separate status today, so the note tells users to confirm a runner advertises matching labels rather than mistaking the wait for a silent failure. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
POST /v1/.../memories and DELETE /v1/.../memories/{memory_key} were
removed from the upstream openapi.json, leaving the nav referencing
endpoints Mintlify can't generate. This drift failed `make lint`
(openapi-drift gate) and the Mintlify Deployment check on main and
every branch off it. Keep GET /v1/.../memories, which still exists
upstream.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Adds a
<Update>for the label-based runner placement capability shipping in usezombie PR #371 (M85_001).What's documented
tags:become itsrequired_tags; a runner claims an agent only when those tags are a subset of the runner's advertised labels. No tags ⇒ any runner (back-compat).UZ-REQ-001.Voice self-audit:
📝 CHANGELOG: 2 bullet(s) · no marketing words · load-bearing facts kept (UZ-REQ-001, bounds, required_tags) · history append-only.🤖 Generated with Claude Code
Greptile Summary
This PR documents the tag-based runner placement feature (usezombie PR #371) by adding a
<Update>block to the changelog and removes two memory API endpoints from thedocs.jsonnavigation.<Update>for Jun 06, 2026 covering tag routing viaSKILL.mdtags:→required_tags, silent-wait behavior with an honest<Note>, andrequired_tagsAPI bounds (1–64 chars per tag, ≤32 per agent,UZ-REQ-001). Terminology, sentence case, and section order align withAGENTS.md.POSTandDELETEmemory endpoints are dropped from the "Agent memories" nav group with no changelog entry and no mention in the PR description; it is unclear whether this reflects a confirmed API removal or is an accidental inclusion in this commit.Confidence Score: 4/5
Safe to merge the changelog entry; the unexplained removal of two memory API endpoints from the navigation needs clarification before merging.
The changelog addition is clean and correct. The concern is in
docs.json:POSTandDELETEmemory endpoints are removed from the nav with no corresponding changelog entry, no PR description mention, and no explanation of whether the underlying API endpoints were actually deprecated. If those endpoints are still live, users lose discoverability; if they were removed, a Breaking changelog entry is required per the project's own companion-docs flow rules.docs.json — the removal of the POST and DELETE memory endpoint nav entries needs confirmation that it is intentional and tied to a live API change.
Important Files Changed
<Update>block documenting tag-based runner placement: tag routing, silent-wait behavior with a<Note>, andrequired_tagsAPI bounds (1–64 chars, ≤32 per agent, UZ-REQ-001). Terminology, sentence case, and section order all match AGENTS.md.POSTandDELETEmemory endpoints from the "Agent memories" nav group without a changelog entry or PR description mention; unclear whether this reflects a live API removal or is an accidental omission.Sequence Diagram
sequenceDiagram participant User participant API participant Agent participant Runner User->>API: POST /zombies (SKILL.md with tags: [gpu, us-east]) API->>Agent: "create agent, derive required_tags=[gpu, us-east]" Note over API: Rejects if tag >64 chars or >32 tags (UZ-REQ-001) Runner->>API: enroll (labels: [gpu, us-east, ...]) API->>Runner: registered Agent->>Runner: claim? (required_tags is subset of runner.labels) Runner-->>Agent: match - run on this runner alt No matching runner enrolled Agent->>Agent: wait silently Note over Agent: No separate status surface today endPrompt To Fix All With AI
Reviews (3): Last reviewed commit: "docs(api): drop dead agent-memories nav ..." | Re-trigger Greptile
Context used: