nexxeln/oc-pstack/reflect

Reflect

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

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.

1. Locate the active transcript

Before fanning out, render this conversation with pstack.transcript({ file: true }) and no sessionID. It writes this session as markdown with [#seq] anchors and returns the path. The pstack.* session tools only see the current project. Never ask for another project's sessions. The transcript header lists subagent sessions. Render one with pstack.transcript({ sessionID, file: true }) when a reviewer needs it. If the render fails, pass a tight digest of the session instead.

2. Spawn three reviewers in parallel

One message, three subagent calls, agent: "general". Reviewers need MCP access to look up the tickets, chat threads, and observability traces the transcript references, and explore has none.

Each reviewer and the synthesizer name a role. Read the roles once with pstack_roles and set the subagent model to that role's value. The user chose these models for pstack, so passing them is an explicit model request. Leave model unset when the value is inherit-parent. If the subagent tool rejects a model, look up the closest model of the same family with opencode.models, use it, and say so. Pick each subagent's reasoning variant for the job, per the Subagents section of the poteto-mode skill (pstack-poteto-mode).

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

3. Synthesize

One subagent call, agent: "general", model from the reflect judgment, divergent, synthesizer role. It spot-verifies citations through MCP, and explore has none. Pass references/synthesizer.md verbatim, with each reviewer's full output inlined where marked. It returns an Accepted / Rejected / Backlog list.

4. Structural enforcement check

Move any Accepted item that a lint rule, script, metadata flag, or runtime check would enforce more reliably to Backlog. See the encode-lessons-in-structure principle skill.

5. Apply

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

File each Backlog item to your team's devex or backlog tracker without waiting. Only the Accepted list waits for approval.

Never edit a pstack skill in the plugin's own skills/ directory. It is generated, and the next update overwrites it. Before approval, show its destination, .opencode/skills/<skill>/ or ~/.config/opencode/skills/<skill>/. After approval, copy the skill folder there and edit the copy. Never open a PR or push.

Follow each approved row's Routing exactly:

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

If your environment ships a SKILL.md validator, run it on every touched skill before declaring done.

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.