Skip to content

feat(skill): self-tagging-first task-journal skill (Track A)#23

Merged
Shahinyanm merged 1 commit into
mainfrom
feat/track-a-self-tagging-skill
Jun 11, 2026
Merged

feat(skill): self-tagging-first task-journal skill (Track A)#23
Shahinyanm merged 1 commit into
mainfrom
feat/track-a-self-tagging-skill

Conversation

@Shahinyanm

Copy link
Copy Markdown
Member

What

Makes explicit agent self-tagging the primary capture path; auto-capture is demoted to a best-effort backstop. Resume packs now come out as Goal → Decisions → Outcome without depending on the classifier.

Changes

  • Rewrote plugin/skills/task-journal/SKILL.md around self-tagging: open a task with a goal, append typed events at the moment of commitment (with alternatives on decisions), task_close with outcome + outcome_tag.
  • Fixed skill ↔ tool drift: dropped the stale evidence_strength reference (it is not an event_add param) and documented the real goal / alternatives / outcome_tag params verified against crates/tj-mcp/src/main.rs.
  • task_create tool description now nudges passing goal (expected-always). No behavior change.
  • README "How it works" leads with self-tagging, reframes auto-capture as a backstop that degrades to heuristic-only without an LLM backend.
  • New regression test crates/tj-mcp/tests/skill_doc.rs: skill present, valid frontmatter, no phantom params, self-tagging framed primary.
  • Plugin minor bump 0.12.0 → 0.13.0 (plugin.json + marketplace.json); CHANGELOG updated.

Verification

  • cargo fmt --all --check clean
  • cargo clippy --workspace --all-targets clean
  • cargo test --workspace 45/45 pass (incl. 3 new skill_doc tests)
  • Dogfood: opened a journal task, logged a decision with alternatives, closed with outcome; pack rendered Goal → Decisions → Outcome (Acceptance A).

🤖 Generated with Claude Code

Rewrite the bundled task-journal skill around agent self-tagging: open a
task with a goal, append typed events at the moment of commitment (with
alternatives on decisions), and task_close with a written outcome +
outcome_tag so resume packs render Goal -> Decisions -> Outcome without
relying on the auto-classifier. Auto-capture is reframed as a best-effort
backstop that degrades to heuristic-only without an LLM backend.

- Drop the stale evidence_strength reference (not an event_add param);
  document the real goal / alternatives / outcome_tag params.
- Tighten task_create MCP tool description to nudge passing goal
  (expected-always). No behavior change.
- README "How it works" leads with self-tagging, demotes auto-capture.
- Add crates/tj-mcp/tests/skill_doc.rs guarding the skill against param
  drift (present, valid frontmatter, no phantom params, self-tagging-first).
- Plugin minor bump: plugin.json + marketplace 0.12.0 -> 0.13.0.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
@Shahinyanm Shahinyanm merged commit 1da81b7 into main Jun 11, 2026
6 of 7 checks passed
@Shahinyanm Shahinyanm deleted the feat/track-a-self-tagging-skill branch June 11, 2026 19:23
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