information-debt-audit
Skill0 starsChanged 19 days agoMITView on GitHub
Audit a set of docs, wikis, dashboards, decks, or folders for information debt — what's outdated, duplicated, conflicting, missing, or ownerless — and produce a prioritized cleanup list. Use when someone says "nobody can find anything," asks to clean up a wiki, Confluence space, Notion workspace, Google Drive, or shared folder, wants a content inventory, or is onboarding and finds conflicting answers.
In this repository33 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- beta-program-managerSkill
- cross-team-dependency-mapSkill
- customer-advisory-board-prepSkill
- data-request-routerSkill
- decision-logSkill
- deprecation-planSkill
- duplicate-request-mergingSkill
- experiment-logSkill
- feature-flag-trackingSkill
- feedback-taggingSkill
- internal-enablement-briefSkill
- kpi-health-checkSkill
- launch-readiness-checklistSkill
- meeting-to-action-converterSkill
- new-pm-onboarding-packSkill
- pricing-packaging-change-logSkill
- process-documentationSkill
- product-metrics-definitionsSkill
- product-ops-contextSkill
- quarterly-planning-prepSkill
- raci-builderSkill
- release-notes-draftingSkill
- request-intake-triageSkill
- retro-facilitationSkill
- ritual-calendarSkill
- roadmap-hygieneSkill
- sprint-review-summarySkill
- support-escalation-patternsSkill
- tool-stack-auditSkill
- vendor-evaluation-scorecardSkill
- voice-of-customer-digestSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 4 are distinctive, and 0 are 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.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 25
If owner or last-updated date is missing, record "None" or "Unknown." That is itself a finding.
- Distinctive: line 39
No owner: nobody is responsible for keeping it current
- Distinctive: line 46
Only flag "Stale" or "Conflicting" with evidence: a date, a newer source, or two items that disagree.
- Distinctive: line 82
Never delete or archive anything yourself.
≈ “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.
Files
Information debt audit
Information debt builds up every time something changes and the docs don't. The cost shows up later as wrong answers, repeated questions, and people trusting the wrong page.
This audit uses Abby Covert's three types of information problems (from How to Make Sense of Any Mess): too much information, not enough information, and not the right information.
Before you start
If .agents/product-ops-context.md exists (or .claude/product-ops-context.md), read it first. Use the team, owners, tools, meetings, and metric definitions it already has, and only ask for what's missing or specific to this task. If it doesn't exist, go ahead with what you're given, and suggest running product-ops-context once so the next skill doesn't have to ask.
Step 1: Inventory
List every item in scope. For each, record what's knowable from the source:
| Item | Type | Owner | Last updated | Audience | Links to / from |
If owner or last-updated date is missing, record "None" or "Unknown." That is itself a finding.
Step 2: Diagnose
Tag each item with every problem that applies.
Too much information
- Duplicate: two or more items cover the same thing
- Bloated: the answer is buried under history, notes, or tangents
- Orphaned: nothing links to it and nobody would find it
Not enough information
- Gap: a question people ask that no item answers
- Stub: started, never finished
- No owner: nobody is responsible for keeping it current
Not the right information
- Stale: describes something that has since changed
- Conflicting: says something different from another item on the same topic
- Wrong audience: written for one group, used by another
Only flag "Stale" or "Conflicting" with evidence: a date, a newer source, or two items that disagree. Quote the conflicting lines side by side.
Step 3: Prioritize
Score each finding on two things:
- Risk: what happens if someone trusts it (wrong customer answer = high, minor confusion = low)
- Reach: how often it's used or found
High risk + high reach goes first.
Step 4: Write the readout
# Information debt audit: [Scope] — [Date]
## Summary
- [N] items reviewed
- Too much: [N] · Not enough: [N] · Not the right: [N]
- Biggest risk: [one sentence]
## Fix first
| Item | Problem | Evidence | Action | Owner |
|---|---|---|---|---|
## Fix next
[Same table]
## Archive or delete
- [Item] — [reason]
## Gaps to fill
- [Question nobody can answer] — [who should own the answer]Actions are one of: Update, Merge into [item], Archive, Assign owner, Create. Every action has a person, not a team.
Never delete or archive anything yourself. Recommend; the owner decides.
Related skills
- process-documentation — rewrite the docs the audit flags
- tool-stack-audit — when the debt is spread across too many tools
- new-pm-onboarding-pack — make sure new hires get current docs
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.

