ke
Skill0 starsChanged 25 days agoMITView on GitHub
Knowledge Extraction — captures and routes knowledge at work boundaries.
In this repository51 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- agent-faultSkill
- apply-capturesSkill
- archgateSkill
- artifactorSkill
- audit-docsSkill
- audit-installationSkill
- author-modeSkill
- bottleneck-pickSkill
- check-secretSkill
- consentSkill
- day-closeSkill
- day-openSkill
- decomposeSkill
- diagnoseSkill
- discovery-sessionSkill
- extendSkill
- fpfSkill
- integration-gateSkill
- iwe-bug-reportSkill
- iwe-platform-redteamSkill
- iwe-restoreSkill
- iwe-rules-reviewSkill
- iwe-updateSkill
- kimi-peer-writerSkill
- local-llmSkill
- month-closeSkill
- org-devSkill
- pack-creatorSkill
- pack-newSkill
- peer-conversationSkill
- personal-guide-renderSkill
- personal-guide-startSkill
- platform-bottleneckSkill
- repo-newSkill
- restore-exocortexSkill
- run-protocolSkill
- setup-wakatimeSkill
- skill-creatorSkill
- strategy-sessionSkill
- skill-idSkill
- thinkSkill
- transcribeSkill
- vdvSkill
- verify-hypothesesSkill
- verifySkill
- week-close-pilotSkill
- week-closeSkill
- wp-newSkill
- w-reflectionSkill
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 2 are distinctive, and 5 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 25
Does not process extraction-reports/*.md (pending-review candidates from agent R2).
- Very specific: line 26
Does not write Pack entities (files DP.*, PD.*, MIM.*).
- Very specific: line 27
Does not launch the extractor agent R2 (runs via launchd every 3h during work hours, invoked as extractor.sh inbox-check).
- Distinctive: line 53
Test HD #29: *"If we replace the vendor, does the statement become false?"*
- Distinctive: line 59
If "Immediately" → record right away
- Very specific: line 60
If "Close" → append an entry to inbox/captures/<YYYY-MM>.md:
- Very specific: line 69
If this is a Distinction — propose a formulation for memory/hard-distinctions.md.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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.
Files
Knowledge Extraction (Capture-to-Pack)
Execute knowledge extraction: $ARGUMENTS
Scope
This Skill does:
- Inline capture at a Session boundary (routing "what → where", announcing "Capture: X → Y").
- Immediate rule recording (1–3 lines) to CLAUDE.md.
- Deferred recording (complex knowledge) — appending to
inbox/captures/<YYYY-MM>.md(routed by event date; rule → Step 3) for the extractor agent. The legacy fileinbox/captures.mddoes not receive new entries (rotation WP-526, 29.08.2026) — treat it as an archive.
This Skill does NOT do:
- Does not process
extraction-reports/*.md(pending-review candidates from agent R2). That is handled by/apply-captures. - Does not write Pack entities (files DP., PD., MIM.*). That is handled by
/apply-capturesafter an R15 decision. Alternatively, R15 can write to the Pack directly via Edit —/apply-capturesis not the only path. - Does not launch the extractor agent R2 (runs via launchd every 3h during work hours, invoked as
extractor.sh inbox-check).
Three knowledge tools (distinctions):
| Tool | Role | When | Output |
/ke (this Skill) | R14 Client / R1 Strategist | Session boundary, inline | Rule in CLAUDE.md / entry in inbox/captures/<YYYY-MM>.md |
extractor.sh inbox-check (agent R2) | R2 Extractor (AI, launchd 3h, work hours) | Automatically | extraction-reports/*.md with status: pending-review |
/apply-captures | R15 Validator | Close when N>0 pending-review | Pack entities + commit + status update |
Detailed VDV cycle map: <governance-repo>/inbox/WP-247-ke-pipeline-vdv.md
Step 1. Identify the knowledge type
| Type | Destination | When | Via KE? |
| Rule for all repos (1–3 lines) | CLAUDE.md (workspace root) | Immediately | No |
| Rule for a single repo (1–3 lines) | <repo>/CLAUDE.md | Immediately | No |
| Domain (architecture, patterns) | Corresponding Pack | Close | Yes |
| Distinction, Method, FM, WP | Corresponding Pack | Close | Yes |
| Implementation (vendor, stack, deployment) | DS docs/ | Close | Yes (KE → DS) |
| Major lesson | memory/<topic>.md | Close | No |
| Post seed | {{GOVERNANCE_REPO}}/drafts/draft-list.md | Close | No |
Step 2. Routing
Test HD #29: "If we replace the vendor, does the statement become false?"
- Yes → implementation (DS)
- No → domain (Pack)
Step 3. Recording
- If "Immediately" → record right away
- If "Close" → append an entry to
inbox/captures/<YYYY-MM>.md:- unambiguous event date in the entry → month of the event;
- event date is absent or ambiguous → month of recording (today), mark in the entry text as
event_date: unknownorevent_date: ambiguous— do not guess inbox/captures.md(legacy) — read-only, do not add new entries there
Step 4. Announce
Output: "Capture: [what] → [where]"
If this is a Distinction — propose a formulation for memory/hard-distinctions.md.
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
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.

