Repository navigation
Conversation
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
pf-guide, a guide to every part of Project Feed for agents on MCP or the pf CLI. Its SKILL.md routes to one reference per area: connection and conventions, posts and versions, media uploads, media review and annotations, tasks, docs, canvases, chat, handoffs, and the CLI.Also rewrites
task-workflow,plan-project,project-status, andwork-triage, and folds inmedia-review-loopfrom #1 with corrections. The old version told agents to subscribe to MCP events, which are off on the hosted server.Every claim was checked against the live staging
tools/listand by calling the tools in a sandbox project. That turned up behavior the docs don't cover:get_taskacceptsPRO-42and42, but most write tools andget_postneed the internal ID.get_postreturns no files.list_mediadoes.get_annotation_framereturns no image.create_posttakes no idempotency key, and unknown arguments are rejected.Behavior evals ran old and new skills on the same 9 tasks against staging fixtures, with Sonnet subagents. The graders check the resulting server state and call logs. The new skills pass 45 of 45 checks and the old ones 41 of 45. The old skills missed verification steps in plans, skipped a post with open feedback in a status report, failed lookups by identifier, and resolved a note on a dropped file without viewing it. The new skills use about 15% more tokens per run.
All client manifests move to 0.4.0. Supersedes #1.