task-organizer
Skill10 starsChanged 6 days agoMITView on GitHub
Mandatory entry point for task folders. Builds the standard four-zone workspace, files every deliverable by provenance, verifies each run with automatic rollback, and keeps the `_broadcast/` index that lets a later session read one summary instead of every original. Fires whenever a task folder is named or attached and work starts inside it.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- JiamingShi-JsPhys/agent-skillstheir own version, first seen Nov 1, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- JiamingShi-JsPhys/agent-skills★ 4same 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 repository22 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .github/copilot-instructions.mdCopilot instructions
- .cursor/rules/chelate-shell.mdcCursor rule
- 3d-modelerSkill
- ai-trace-removerSkill
- bibtex-normalizerSkill
- capstone-labmateSkill
- conflict-detectorSkill
- data-profilerSkill
- document-evolution-trackerSkill
- formula-validatorSkill
- fulltext-extractorSkill
- latex-exporterSkill
- literature-retrieverSkill
- longitudinal-analysisSkill
- ml-auditorSkill
- multi-dim-managerSkill
- quantitative-validatorSkill
- reference-auditorSkill
- sci-visualizerSkill
- security-guardSkill
- video-trackerSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 2 are distinctive, and 2 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.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 29
Run this exact form — do not read the source tree:
- Very specific: line 40
Every run writes its gate to <WS>/process/gates/task-organizer_<ts>.json.
- Distinctive: line 51
No workspace has been chosen — do not organize an unnamed folder.
- Very specific: line 73
Never move user material: Only init relocates pre-existing content, and only after the user agrees.
≈ “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
task-organizer
Builds and maintains the workspace skeleton — material/initial/,
material/downloaded/, process/, results/ and _broadcast/ — files each
deliverable by provenance, and verifies each run. It never edits the content of
your documents; it only decides where files belong.
- Stage:
workspace· Sandbox: no · Dependencies: stdlib only (openpyxladds the xlsx index,PyYAMLthe YAML rules)
Quick Execution
Run this exact form — do not read the source tree:
python run.py --root <WS> --action doctordoctoris read-only and always safe: it reportshealthyorincompleteand writes nothing. Run it before asking the user anything.initcreates the whole skeleton and lifts pre-existing content intomaterial/initial/. It moves files, so propose it once and wait for a yes when the folder was merely referenced.- Every run writes its gate to
<WS>/process/gates/task-organizer_<ts>.json.
When to use
- At session start on a known task folder — run
doctorfirst. - At task start on a new or empty folder, or on an explicit "set this up" request — run
init. - At final delivery — run
organize, thenbroadcast(the index is mandatory, not optional). - A fetched file must be declared provenance —
mark-downloaded, somaterial/downloaded/means "the assistant fetched it" rather than a guess from the filename.
When NOT to use
- No workspace has been chosen — do not organize an unnamed folder.
- The user declined the
initproposal — carry on as asked, at most one proposal per folder. - You were asked to summarize or describe this skill — that is a failure mode, not a result.
Actions
--action | Does | Extra arguments |
doctor | Read-only structure check | — |
init | Build the four zones; pre-existing content moves to material/initial/ | — |
organize | Full pipeline: file, validate (C1–C8), verify, rollback on failure | --mode full|quick|incremental (after --) |
broadcast | Rebuild the read-only _broadcast/INDEX.md | --ai, --max-files N, --keywords K |
index | Alias of broadcast | same as broadcast |
snapshot | Hash snapshot of the current tree | — |
dedupe | Report duplicate files by content hash | — |
mark-downloaded | Record a fetched file's source URL | <file> --url <source-url> |
mark-user | Declare a file as user material | <file> |
undo / redo | Reverse or replay a verified run | <history_file> |
Hard constraints
- Act, do not summarize. A trigger under "When to use" means run the tool and report what it did.
- Never move user material. Only
initrelocates pre-existing content, and only after the user agrees. - A failed verification has already rolled back. Report it; do not blindly re-run.
Outputs
<WS>/process/gates/task-organizer_<ts>.json— the gate result.<WS>/_broadcast/INDEX.md— the context-compression index.<WS>/process/history/— the reverse-able run history used byundo.
Failure handling
verification: failed— the tool restored the before-state already; report that and stop.- A user-material name clash is left in place and reported, never overwritten.
- Hidden files, empty folders and symlinks are skipped by design.
- Exit
0success,2usage or validation error.
Limitations
- Filing is by provenance and name pattern, not by semantics.
broadcastindexes what is on disk; it cannot judge whether a file is correct.
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.