Skip to content
Agentmart
Submit a tool

Remember

Memory hygiene workflow to classify, deduplicate, and promote durable knowledge across local memory layers.

Skill by lordmacu

IndexedApache-2.0★ 8
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/lordmacu-remember

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Memory hygiene workflow to classify, deduplicate, and promote durable knowledge across local memory layers.

Use cases: productivity

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/remember

What's inside

Skill name
Remember
Description
Memory hygiene workflow to classify, deduplicate, and promote durable knowledge across local memory layers.
License
Apache-2.0
Instructions
76 lines in SKILL.md · view on GitHub

Agent Skills spec issues

  • name: only lowercase letters, digits and single hyphens allowed
  • name: "Remember" does not match folder "remember"

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Remember

Review memory-like artifacts and propose what should be kept, promoted, merged,
or removed. This keeps guidance accurate as the project evolves.

## Use when

- Memory instructions are drifting or conflicting.
- Team wants to clean stale guidance from old sessions.
- You need to promote stable patterns into durable docs.

## Do not use when

- There is no meaningful memory/context to review.
- User only needs a one-off answer (no memory maintenance).
- You are asked to auto-edit memory layers without confirmation.

## Input contract

- `review_scope` (optional): `workspace_docs`, `session_logs`, `both` (default `both`).
- `apply_changes` (optional): default `false` (proposal mode first).
- `priority` (optional): `conflicts`, `duplicates`, `promotions`, `all` (default `all`).
- `target_files` (optional): explicit files to inspect first.

## Parsing rules (priority order)

1. Use explicit structured args when provided.
2. Else infer from user phrasing:
   - "clean memory", "dedupe instructions" -> `priority=duplicates,conflicts`
   - "promote stable conventions" -> `priority=promotions`
   - "just review, don't edit" -> `apply_changes=false`
3. Defaults:
   - `review_scope=both`
   - `apply_changes=false`
   - `priority=all`

## Canonical workflow

1. **Collect sources**
   - Workspace memory docs (for example: `IDENTITY.md`, `SOUL.md`, `MEMORY.md`,
     `USER.md`, `AGENTS.md`) and session evidence when available.
2. **Classify entries**
   - durable convention, temporary note, duplicate, outdated, conflicting, ambiguous.
3. **Generate proposals**
   - promotions (move to durable location), cleanup (remove/merge duplicates),
     conflict resolution (which version wins + rationale).
4. **Present grouped report**
   - `Promotions`, `Cleanup`, `Conflicts`, `Ambiguous`, `No action`.
5. **Apply only with approval**
   - If `apply_changes=true`, execute approved changes and re-validate consistency.

## Guardrails

- Default to proposal mode; do not mutate without explicit approval.
- Never delete content that has unresolved ambiguity.
- Prefer reversible edits and clear provenance notes.
- Keep sensitive values out of memory docs; redact before proposing promotion.
- State uncertainty explicitly when evidence is insufficient.

## Output format

Always return:

- `status`: `proposed` | `applied` | `needs_user_input`
- `review_scope`: effective scope
- `promotions`: list with destination + rationale
- `cleanup`: duplicates/outdated removals or merges
- `conflicts`: contradiction list + recommended resolution
- `ambiguous`: items requiring user decision
- `next_action`: required unless `status=applied`

## Example

- `review_scope="both" priority="all" apply_changes=false`

From lordmacu/nexo-rs, Apache-2.0.

More from lordmacu/nexo-rs

Signals

Stars
8
Forks
4
Open issues
0
Last commit (this folder)
Apr 29, 2026 (6 months ago)
Last release
May 11, 2026 (5 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
Apache-2.0
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from lordmacu

Cite this page

Remember skill. Agent Mart. https://agentmart.dev/skills/lordmacu-remember (last updated 2026-10-24)