retention
Skill2 starsChanged 50 days agoMITView on GitHub
Act on the storage clock — transcripts past their retention date. Use on /retention, when graph-lint reports an overdue file, or during a periodic data review.
6 other projects use a version of this.
Adopted by
Most recently taken up by
- jestermaxdelta/roblox-building-skilltheir own version, first seen Nov 16, 2026
- whawkinsiv/solo-founder-skillstheir own version, first seen Oct 25, 2026
- whawkinsiv/claude-code-skillstheir own version, first seen Oct 17, 2026
- akachandan1/proofstacktheir own version, first seen Oct 11, 2026
- Infinite-Labs-AI/infinite-skillstheir own version, first seen Jul 2, 2026
- ORBWEVA/founder-opstheir own version, first seen Mar 10, 2026
Other versions6 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 6
- whawkinsiv/solo-founder-skills★ 260same name
- whawkinsiv/claude-code-skills★ 258same name
- Infinite-Labs-AI/infinite-skills★ 43same name
- jestermaxdelta/roblox-building-skill★ 1same name
- akachandan1/proofstack★ 0same name
- ORBWEVA/founder-ops★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository31 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- affinitySkill
- ai-personaSkill
- analytics-syncSkill
- backlogSkill
- catalogSkill
- cold-startSkill
- contradictionsSkill
- demo-dataSkill
- dovetail-syncSkill
- exportSkill
- extract-findingsSkill
- feature-panelSkill
- forget-participantSkill
- graph-lintSkill
- interview-guideSkill
- journey-mapSkill
- opportunity-treeSkill
- persona-avatarSkill
- persona-querySkill
- persona-talkSkill
- persona-voiceSkill
- persona-workshopSkill
- prdSkill
- prioritizeSkill
- researcherSkill
- source-syncSkill
- subject-accessSkill
- welcomeSkill
The file
We split this skill into the rules it gives an agent
2 rules. 0 are common, 1 is distinctive, and 1 is very specific to this project.
0
Common
Most of the 6 edited versions we read keep these.
1
Distinctive
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 59
If the material can't be de-identified without gutting it — a small sample where the story itself names the person, a session about a named …
- Very specific: line 81
If the consent said "we keep recordings 12 months", that is the date.
“Common” is measured: kept by at least 6 in 10 of the 6 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.
Files
retention — the clock, and why it usually ends in a state change
Storage limitation (GDPR Art. 5(1)(e)) says personal data may not be kept longer than necessary. It does not say "delete your research" — and treating the two as the same is how teams lose their evidence base for no compliance gain.
The insight this skill is built on: anonymous data is outside the GDPR entirely. So the natural end of the clock is not deletion, it is graduation.
raw → scrubbed → key destroyed (anonymous) → [only if needed] deleted
↑ clock running ↑ clock stopsTrigger
/retention ← what is due, what is close, what to do
/retention --review ← walk every transcript, not only overdue onesgraph-lint surfaces overdue files too and points here.
Workflow
1. What is due
python3 scripts/graph_lint.py # [retention] lines: overdue, and due within 30 daysA file only has a clock if it has retention_until: and is still personal data
(key_location: anything but destroyed). Files without the field aren't
violations — they're unrecorded. Offer to set a date, don't nag.
2. Propose the state change first
For each overdue transcript, in this order:
- Finish de-identification — is the scrub good enough that nobody could
identify this person from the remaining text? Check the whole thing, not just
names: unique phrasing, employer, city plus role plus age, an anecdote only
one person could tell.
anonymization: manual_verifiedmeans a human confirmed this; nothing else does. - Destroy the key — the linking list, the recording, the calendar invite,
the recruiting-platform record, the incentive payment trail. This happens
outside this repo, in the systems that hold them; the skill can only set
key_location: destroyedto record that it was done. Ask, don't assume. - Once both are true, the file is anonymous: set
key_location: destroyed, dropretention_until(or leave it as history), and the clock stops for good.
3. Delete only when the text still identifies
If the material can't be de-identified without gutting it — a small sample where
the story itself names the person, a session about a named product decision —
then deletion is the right answer. Hand off to /forget-participant <CODE>,
which does the cascade and the re-grading properly.
4. Report
Show: what graduated (clock stopped), what was deleted, what the user deferred
and until when. Then run graph_lint to confirm the graph is still consistent.
Rules
- Never delete a file on your own. Propose; the researcher decides. This skill
edits frontmatter after approval and hands deletion to
/forget-participant. - De-identification before deletion, always. Deleting a transcript that could simply have been anonymised destroys research for nothing.
- Retention ≠ freshness. The 3-month freshness rule asks is this finding still true?; retention asks may we still hold this? A stale finding may be perfectly lawful to keep; a fresh one may be past its retention date. Never merge the two clocks, and never let a UI filter that hides old items be described as deletion.
- The clock comes from your promise, not from a statute. GDPR sets no number.
If the consent said "we keep recordings 12 months", that is the date. Read
consent_scopeandconsent_refbefore proposing anything. - Demo files have no clock — fictional participants, nothing to expire.
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.

