loose-ends
Skill7 starsChanged 3 months agoMITView on GitHub
Surface unaddressed threads from the conversation — bugs mentioned but not fixed, deferred decisions, dropped nits, cross-workstream dependencies, unverified claims, and promised-but-not-done items — as a tight, skimmable report. Reports; never edits code. Use when the user says "loose ends", "any loose ends", "what did we leave hanging", "what didn't we get to", "what got dropped", "open threads", "unaddressed items", "what did we punt on", or "/loose-ends".
5 other projects use a version of this.
Adopted by
Most recently taken up by
- nmamano/isomuxtheir own version, first seen Nov 1, 2026
- TucsenRoman/claude-skillstheir own version, first seen Oct 20, 2026
- ai-ecoverse/skillstheir own version, first seen Oct 2, 2026
- paradite/skillstheir own version, first seen Aug 22, 2026
- noron12234/loose-endstheir own version, first seen Jul 29, 2026
Other versions5 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 5
- nmamano/isomux★ 43same name
- ai-ecoverse/skills★ 1same name
- noron12234/loose-ends★ 1same name
- paradite/skills★ 1same name
- TucsenRoman/claude-skills★ 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 repository51 other files it gives its agents.
- AGENTS.mdAGENTS.md
- skills/optimize-permissions/references/agents.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- acid-tripSkill
- audit-aws-costsSkill
- audit-memoriesSkill
- broSkill
- clean-ai-slopSkill
- cleanup-computerSkill
- code-tourSkill
- copywritingSkill
- decomposeSkill
- eli5Skill
- find-docsSkill
- gemini-deep-researchSkill
- investigate-repoSkill
- laconicSkill
- memory-insightsSkill
- nano-banana-generatorSkill
- optimize-permissionsSkill
- optimize-skill-activationSkill
- projectSkill
- prSkill
- quality-chaos-monkeySkill
- quality-cli-agent-friendly-auditSkill
- quality-dead-code-analyzerSkill
- quality-docs-updateSkill
- quality-perf-reviewSkill
- quality-project-healthSkill
- quality-worktree-setupSkill
- record-gifSkill
- reflect-on-conversationSkill
- review-codeSkill
- safety-dotenv-guardSkill
- safety-git-reset-guardSkill
- safety-op-credsSkill
- safety-rm-rf-guardSkill
- security-bumblebeeSkill
- security-gitleaksSkill
- security-openssfSkill
- security-scfwSkill
- security-snykSkill
- security-socketSkill
- ste100Skill
- storage-cleanupSkill
- testSkill
- transcribe-youtubeSkill
- travel-photo-stampSkill
- trip-plannerSkill
- zoomSkill
- __SKILL_NAME__Skill
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 4 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 29
Read the whole conversation.
- Very specific: line 72
Use [unsure] only when useful to show that the item could not be confirmed.
“Common” is measured: kept by at least 6 in 10 of the 4 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
/loose-ends — Surface Unaddressed Threads From the Conversation
You are auditing the conversation itself, not a diff. Read back over the session and surface every loose end: anything raised that could plausibly need action later, where the conversation moved on without resolving it. Output a concise, scannable list. This is a reporting step — it never edits code and proposes no fixes.
The source of truth is the transcript, not git diff. A loose end may have no
code footprint at all (a deferred decision, a dependency on another PR). Do not
restrict yourself to the diff.
When to use
- "loose ends" / "any loose ends" / "what did we leave hanging"
- "what didn't we get to" / "what got dropped"
- "open threads" / "unaddressed items" / "what did we punt on"
- "/loose-ends"
Scan
Read the whole conversation. Look for:
- defects mentioned but not fixed;
- choices deferred or made implicitly without surfacing the tradeoff;
- acknowledged shortcuts, dropped nits, and TODOs;
- dependencies or conflicts with other work that were not acted on or tracked;
- assumptions that the conversation relied on without checking;
- commitments that the assistant or user made but did not complete.
These are search prompts, not output categories. A candidate belongs in the report only when it is both unresolved and plausibly action-worthy.
Re-read the later conversation first so you do not report something quietly
handled afterward. Verify against the current code only when the claim is
concrete, local, and cheap to check. If it would require a broader investigation,
keep the item only when it still matters and mark it [unsure].
Skip rhetorical asides, rejected ideas, and work already captured in a ticket or clearly intentional tracked TODO. Prefer a short, high-signal list. If nothing qualifies, say so plainly.
Output format
Return one ranked list, most actionable first. Keep each item to one sentence:
what remains unresolved, plus where when a useful file:line, commit, PR, or
ticket exists. Do not propose fixes.
## Loose ends
- Empty-input path still dereferences `value.trim()` before the null check —
flagged mid-session, never patched. `src/ui/Input.tsx:42`
- [unsure] Timezone handling in the export may double-apply the offset — raised,
not confirmed either way.
- Chose in-memory caching without surfacing the Redis option to you — worth a
yes/no before this ships.
- Depends on the auth refactor in PR #214 landing first; not tracked anywhere.
- Said a regression test would be added for the retry fix — none written.
5 loose ends.Rules:
- Use
[unsure]only when useful to show that the item could not be confirmed. - Stay concrete: "Error handling could be better" is noise; "The
parse()path swallows aSyntaxErrorwe said we'd surface" is a loose end. - End with
N loose ends.Do not add categories or an area count. - Report only. Never edit code or open issues as part of this skill.
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.

