Reflect

Claude Code: read Cursor terms in this skill through claude-code/runtime.md (tools, subagents, models, transcripts).

Mine the current conversation for durable learnings, then route them into skill edits.

When to invoke

Invoke when the user says "reflect" or "/reflect". Skip when the conversation is trivial, off-topic, or already covered by an existing skill the parent followed correctly. One-offs are not learnings.

Process

1. Locate the active transcript

The parent finds its own transcript file before fanning out. Claude Code keeps it under ~/.claude/projects/<project>/ (see pstack's claude-code/runtime.md). Use that path. Do not glob across ~/.claude/projects/*/. That crosses workspace boundaries and reads private chats from unrelated projects.

ls -t ~/.claude/projects/<project>/*.jsonl ~/.claude/projects/<project>/*/*.jsonl ~/.claude/projects/<project>/*/subagents/*.jsonl 2>/dev/null | head -10

Three transcript layouts: legacy flat (<id>.jsonl), current nested (<id>/<id>.jsonl), and subagent (<parent>/subagents/<child>.jsonl).

For each candidate, read the first JSONL line and check that message.content[0].text contains the conversation's opening user prompt. Take the matching path. If no path resolves, write a tight digest of the session and pass that instead.

2. Spawn three reviewers in parallel

One message, three Agent calls, subagent_type: general-purpose, with explicit model and effort set as below, full tool access (general-purpose, MCP included). Reviewers need MCP access for context lookups (tickets, chat threads, observability traces referenced in the transcript). Readonly strips MCPs.

Resolve each seat's explicit model and effort from delegation-kit's live routing table when its Delegation routing (live section is in context; otherwise use pstack-models.md and the defaults below. Map auto, inherit-parent, and inherit to model: sonnet, effort: medium. Always pass both parameters to Agent, including retries and resumes. For fan-outs of 3+ use sonnet/haiku seats, never opus. The main session owns final synthesis; any separately prescribed synthesizer runs alone after the fan-out.

Pass each template verbatim, substituting the transcript path or digest where marked. Reviewers return findings in the Agent response body.

3. Synthesize

One Agent call, subagent_type: general-purpose, with explicit model and effort from the reflect judgment, divergent, synthesizer line (single-seat default opus / medium, after all reviewers finish), full tool access (general-purpose, MCP included). The synthesizer's quality check includes spot-verifying citations, which can require MCP access. Readonly strips MCPs. Use references/synthesizer.md verbatim, with each reviewer's full output inlined where marked. The synthesizer returns a structured Accepted / Rejected / Backlog list.

4. Structural enforcement check

Sanity-check the synthesizer's Accepted list. For any item that would be enforced more reliably by a lint rule, script, metadata flag, or runtime check, move it from Accepted to Backlog. See the encode-lessons-in-structure principle skill.

5. Apply

Before applying any Accepted edit, present the synthesizer's full Accepted/Rejected/Backlog output to the user and wait for explicit approval. The user picks which subset to apply and may redirect routings. Skill changes affect every future agent in the org. Do not auto-apply.

Backlog items file to whatever devex / backlog tracker your team uses automatically. Only the Accepted list waits for approval.

For each approved Accepted item, follow the Routing field exactly:

• Trivial existing-skill edit (a one-line bullet, a tightened sentence, a stale fact corrected): parent does directly.
• Substantive existing-skill edit (a new section, a new pattern table, more than ~10 lines): hand to the skill-creator skill (see pstack's claude-code/runtime.md) and run its draft / test / iterate loop.
• tune description: <skill path> (the skill exists but didn't trigger when it should have): hand to create-skill and run its description-optimization loop.
• new skill via create-skill: <kebab-name>: hand creation to create-skill. Do not invent the shape ad hoc.

If your environment ships a SKILL.md validator, run it on every touched skill before declaring done. Skip this step if it doesn't.

6. Summarize for the user

Short list, no preamble:

• Edits applied: <skill path>. What changed, one line each.
• New skills created: <skill path>. One line each (rare).
• Backlog filed to the devex tracker: <issue title> (<tags>). One line each.
• Dropped: one line per rejected finding + reason from the synthesizer.