agentleFS
niksavisniksavis/basicly/.agents/skills/session-finish/SKILL.md

session-finish

Skill2 starsChanged 6 days agoMITView on GitHub

Summarizes what changed and what is still open. Use to wrap up a session or run a retro. Write only new recovery facts.

1 other project uses a version of this, 1 of them edited.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 4 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Never invent a retro finding: When nothing concrete happened, skip the retro and say so.

    line 20
  2. Very specific:

    Run basicly tracker show <id> --view current on the root record first.

    line 26
  3. Distinctive:

    Use usage data only when it answers the current question.

    line 30
  4. Distinctive:

    Run it after a refused change or a corrected mistake whose cause remains unresolved.

    line 34
  5. Distinctive:

    Before closing the session, read clean handoff for tree cleanup, record closure, process ownership and the root handover note.

    line 40

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md
<!-- Generated by `basicly skills-build` from skill.yaml. Do not edit; edit the source. -->

Session Finish

Finish landed work with current tracker facts, necessary recovery state and a clean repository. Usage reports and retrospectives are conditional, not phases that require new agents or commits. Use work-tracker for history and record closure and conventional-commits for commits. Use harness-loop only for explicitly loop-managed work.

Rules

  • Keep recovery facts in the existing ledger. basicly tracker show <id> --view current preserves current fields, clarifications and latest attributed handoffs. Retrieve omitted history with --view history --snapshot <snapshot_identity> for unresolved or conflicting facts. The view is recovery guidance, not a complete contract.
  • Write only new recovery facts. Preserve decisions, changed assumptions, uncertainty, blockers, evidence references, owned process handles and the next action when they cannot be derived. Mark these notes with [new facts] immediately after the handover marker so the current view retains each addition. Do not repeat current card fields, Git history or unchanged handoffs on routine transitions.
  • Verify a recorded process handle before replacing it. A polling timeout is not proof that a process stopped. Resume or inspect the same handle through its owning host or tool; do not launch a duplicate. A PID is local to its host and process namespace. If that owner cannot be reached, report the process state as unknown instead of stopped.
  • Stop task-owned temporary processes before handover. Preserve services the user explicitly asked to keep running; record their URLs and ownership. Never stop another session's process.
  • Keep every durable artifact in the repository. The host cleans the scratchpad without warning.
  • Write scratch files to a subdirectory unique to your lane. A concurrent lane can overwrite a file at a shared path.
  • Apply a retro fix within the approved scope. Record other recommendations for approval; do not ask again for work the owner already authorized.
  • Never invent a retro finding. When nothing concrete happened, skip the retro and say so.
  • Never reset or delete the usage counters. They accumulate across sessions.
  • In a long run, report the elapsed time, the tokens spent where the host shows them, and the items left. Report at each milestone, not only at the end.

0. Read the root record

Run basicly tracker show <id> --view current on the root record first. Write the handover from what it already holds.

1. Usage statistics

Use usage data only when it answers the current question. For the report commands and how to read them, read usage reports.

2. Retro

  • Run it after a refused change or a corrected mistake whose cause remains unresolved. Reuse an existing diagnosis and fix; do not repeat the investigation for the handover.
  • Propose the exact fragment, skill or hook change that would have refused it.
  • Record an environment, timing or platform trap as one bullet in a quirks fragment: the trap and the avoidance.

3. Clean handoff

Before closing the session, read clean handoff for tree cleanup, record closure, process ownership and the root handover note.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.