Self-correcting memory for LLM agents. A Claude Code plugin that learns from sessions, surfaces relevant memories, measures whether they're actually followed, and fixes the ones that aren't.
This repository also configures its own agents. See what engram tells them →
Skill Claude CodeCodex
Use when an engram vault holds pending offers awaiting review — engram query's payload shows pendingoffers: true, an engram update notice names pending offers, a write-path warning nudge fires after engram learn/amend/resituate, or a child's offer or a pulled-down parent note is pending. Also use when explicitly asked…
Skill Claude CodeCodex
Use after work begun with recall, or any work requiring more than one tool call or non-trivial thought. Use immediately for explicit save requests ("remember this", "save that", "note for next time"), and the moment a review, failing check, or user correction confirms a lesson; capture it before applying the fix. Also…
Skill Claude CodeCodex
Use when the user asks you to take a task end-to-end. Triggers on the explicit slash form /please and on natural-language phrasings of the same intent: "please do X", "please work through X", "please take care of X", "work through X end-to-end", "take this end-to-end", "drive X to completion", "see this through". An…
Skill Claude CodeCodex
Use after any user request that might entail more than a single tool call or anything more than quick, shallow thinking. This surfaces relevant memories that are VITAL to recall for a good user experience and a greater chance at first-pass success for the user's request.
Skill Claude CodeCodex
Use when you are about to dispatch a subagent and must decide its agent type, model, and effort level. Triggers on any delegation decision, and when you recognize a unit is too large for one focused agent and needs decomposition before dispatch.
Skill Claude CodeCodex
Executes a vault write handed off by another skill (recall, learn): composes the engram command from the provided fields, runs it, verifies the result, and reports the written note path. Requires a handoff — do not fire on your own judgment that something is worth remembering.
Skill Claude CodeCodex
Use when registering or adding a new sensor type to the telemetry system's registry, codegen, migrations, and changelog.
Skill Claude CodeCodex
Use when narrowing or rewriting a .gitignore pattern so that some previously-ignored files become trackable, before staging or committing that change. Symptoms/keywords — gitignore anchoring, git check-ignore, middle-slash patterns, testdata/ narrowing, newly-visible or newly-untracked files, git add -A, git add .
Skill Claude Code
Use when a trap/eval gate returns RED on the current branch and you need to determine whether the regression is pre-existing or caused by the change under review, before applying any prescribed fix. Keywords — gate regression, RED verdict, pre-existing failure, change attribution, gate.py, crowdedgate.py, bisecting…
Skill Claude CodeCodex
Use when rewriting git history with filter-branch or similar all-refs rewrites, especially before force-pushing with a backup branch present. Keywords — backup branch sweep, refs/original cleanup, force-with-lease, tracking refs refresh, pre-rewrite recovery.
Skill Claude CodeCodex
Use when implementing any feature or bugfix, before writing implementation code.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: