agentleFS
markmhendricksonmarkmhendrickson/neotoma/.claude/skills/end/SKILL.md

end

Skill38 starsChanged 2 months agoMITView on GitHub

Session-end audit. Surfaces remaining work to complete or track, verifies all data intended for Neotoma storage from this session is actually stored, and offers to file what's missing. Invoke at the natural close of a working session, before context is lost.

11 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use update.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • AnthropicSkill

    update

    Sync tasks and refresh memory from your current activity. Use when pulling new assignments from your project tracker into TASKS.md, triaging stale or overdue tasks, filling memory gaps for unknown people or projects, or running a comprehensive scan to catch todos buried in chat and email.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

9 rules. 0 are common, 3 are distinctive, and 6 are very specific to this project.

  • 0

    Common

    Most of the 10 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 6

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If the conversation itself has not been persisted as a conversation + dual conversation_message shape end-to-end, invoke store-neotoma to …

    line 56
  2. Very specific:

    Write any memory files per the auto-memory protocol and update MEMORY.md index.

    line 57
  3. Very specific:

    If do-now items exist and the user requested execution, perform them inline.

    line 58
  4. Distinctive:

    MUST NOT execute storage or task filing without user confirmation of the Phase 3 plan.

    line 80
  5. Distinctive:

    MUST NOT skip the storage audit even if the user only asks about remaining work, and vice versa.

    line 81
  6. Distinctive:

    MUST classify every surfaced item.

    line 82
  7. Very specific:

    MUST check existing Neotoma state via retrieval before declaring an item "not stored."

    line 83
  8. Very specific:

    MUST defer chat-transcript storage to store-neotoma, not re-implement it.

    line 84
  9. Very specific:

    MUST strip PII from any filed issues per feedback_pii_in_issues memory.

    line 85

“Common” is measured: kept by at least 6 in 10 of the 10 edited versions of this file we could read. “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

end

Purpose

Run a session-close audit so nothing intended for follow-up or for Neotoma storage falls through the cracks. Distinct from the primitive store MCP tool call (single per-record persistence) and /store-neotoma (full chat-transcript persistence): /end is the meta-step that decides which of those to invoke, and what remaining tasks need to be tracked.

Scope

Applies once per session, at user request. Does not modify code. Does file Neotoma entities (tasks, issues, plan updates) and may delegate to store-neotoma for chat persistence.

Phase 1: Remaining-work audit

Scan the current conversation and produce a structured list under these headings. List items only; do not write to disk yet.

  1. Open work — TODOs the assistant introduced, partial implementations, files modified but not verified, tests not run, lint/type-check skipped, PRs/commits not made.
  2. Decisions or proposals not acted on — recommendations the user accepted that have not been executed; designs sketched but not implemented.
  3. Trackable follow-ups — bugs noticed in passing, refactors deferred, documentation drift spotted, dependencies needing updates.
  4. External obligations — anything waiting on CI, a remote agent, a scheduled task, or a third party.
  5. User feedback or preferences expressed this session — candidates for memory (feedback/user/project/reference).

For each item, classify: do-now (small enough to finish in this session), track (file as task/issue/plan entry), or drop (acknowledged, no action).

Phase 2: Neotoma storage audit

Determine what should be in Neotoma from this session and what already is.

  1. Per-turn lifecycle compliance — confirm each turn this session followed the Neotoma turn lifecycle (user message + assistant message stored, PART_OF + REFERS_TO edges). If any turn was skipped (which is forbidden), list it.
  2. Substantive entities surfaced — list every concrete entity discussed or produced this session: plans, decisions, skills, rules, bugs, contacts, transactions, events, code artifacts, etc. For each, check via retrieve_entity_by_identifier or retrieve_entities whether it is already stored.
  3. Files or attachments — any files the user pasted, screenshots, transcripts, or external URLs fetched. Check whether each has a corresponding source_id / content-addressed source row.
  4. Memory-worthy facts — items that should be written to the auto-memory directory (~/.claude/projects/.../memory/) per the auto-memory protocol.

Produce a table with columns: item, kind (entity / source / memory), stored? (yes/no/partial), action (store via the store MCP tool / store via store-neotoma / write memory file / skip).

Phase 3: Confirmation and execution

Present Phases 1 and 2 to the user as a single end-of-session report with two sections: "Remaining work" and "Storage gaps." End with:

  • A concrete list of proposed actions (e.g., "file 3 tasks, store 2 entities, invoke /store-neotoma, write 1 memory file").
  • Request confirmation, or accept revisions.

On confirmation:

  1. File task entities for each track item from Phase 1 via the store MCP tool with appropriate REFERS_TO edges to relevant entities.
  2. Store missing entities or sources flagged in Phase 2 via the store MCP tool directly.
  3. If the conversation itself has not been persisted as a conversation + dual conversation_message shape end-to-end, invoke store-neotoma to do the full transcript sweep.
  4. Write any memory files per the auto-memory protocol and update MEMORY.md index.
  5. If do-now items exist and the user requested execution, perform them inline; otherwise file them as tasks.

Phase 4: Final summary

Report:

  • Remaining-work items filed (with task entity IDs).
  • Entities/sources stored (with entity IDs).
  • Memory files written.
  • Anything dropped or deferred, with one-line reason.
  • Whether store-neotoma was invoked.

Render the mandatory 🧠 Neotoma turn report covering all entities created in Phases 3–4.

Relationship to other skills

  • store (MCP tool) — per-entity/per-file store primitive. /end calls it directly for each gap identified in Phase 2.
  • store-neotoma — full chat-transcript persistence with preview/confirm. /end delegates to it when the conversation as a whole is not yet stored.
  • Redundancy note: the primitive store MCP tool and store-neotoma are not redundant — direct store calls are per-record and per-call; store-neotoma is the bulk session-end transcript sweep with preview.

Constraints

  • MUST NOT execute storage or task filing without user confirmation of the Phase 3 plan.
  • MUST NOT skip the storage audit even if the user only asks about remaining work, and vice versa.
  • MUST classify every surfaced item; no unclassified entries.
  • MUST check existing Neotoma state via retrieval before declaring an item "not stored."
  • MUST defer chat-transcript storage to store-neotoma, not re-implement it.
  • MUST strip PII from any filed issues per feedback_pii_in_issues memory.

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.