unfinished-work-tracking
Skill0 starsChanged 3 days agoMITView on GitHub
Unfinished work tracking standard. Requires recording requirements, defects, omissions, and items pending verification that are discovered during development but not yet resolved.
In this repository17 other files it gives its agents.
- .cursor/rules/engineering-context.mdcCursor rule
- change-report-deliverySkill
- code-asset-inventorySkill
- data-audit-reproducibilitySkill
- data-first-executionSkill
- documentation-structureSkill
- engineering-context-completenessSkill
- engineering-transparencySkill
- explicit-instruction-fidelitySkill
- interface-contractSkill
- language-expressionSkill
- production-loggingSkill
- repository-publication-hygieneSkill
- task-execution-governanceSkill
- technical-documentationSkill
- testing-verificationSkill
- user-intent-prioritySkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
13 rules. About 0 are common, 13 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈13
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 20
Only part of a feature was completed.
- Distinctive: line 40
If the project has no unified recording location, use TODO.md.
- Distinctive: line 52
Do not build an unnecessarily complex task management system.
- Distinctive: line 59
Check whether the current task involves any of them.
- Distinctive: line 61
If they affect the correctness of the current task, resolve them first.
- Distinctive: line 65
When a new item is discovered:
- Distinctive: line 67
If it can be safely completed: complete it directly.
- Distinctive: line 68
If it requires additional work: record it.
- Distinctive: line 69
If it is temporarily impossible: record the reason.
- Distinctive: line 70
If an existing item has been resolved: update its status.
- Distinctive: line 71
If an old record turns out to be inaccurate: correct the record.
- Distinctive: line 88
Only work that was actually completed may be marked as completed.
- Distinctive: line 98
Not having told the user does not mean there are no outstanding items.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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
Unfinished Work Tracking Standard
I. Core Principle
Any relevant work discovered during execution but not actually completed must be recorded.
Unfinished information must not be discarded merely because the current task ended, the context switched, or the user did not ask about it.
II. Situations That Require Recording
Include:
- An explicit requirement that has not been implemented.
- Only part of a feature was completed.
- A gap was discovered in existing functionality.
- A problem was discovered that needs fixing.
- An interface was discovered that needs adjustment.
- Documentation was discovered to be out of sync.
- A matter was discovered that needs further investigation.
- Something temporarily cannot be completed due to environment or dependency issues.
- Something believed to be implemented was found, upon inspection, to actually not be implemented.
III. Situations That Do Not Require Recording
- Random thoughts unrelated to the current task.
- Problems that have actually been completed.
- Speculation with no clear action value.
- Duplicate items already recorded with no new information.
IV. Record Format
By default, use the project's existing TODO file.
If the project has no unified recording location, use TODO.md.
Format:
- Outstanding item
- Reason: why it needs to be handled.
- Impact: which function it affects.
- Related files: which files are involved.
- Current status: why it is not completed.
For complex items, add necessary reproduction steps and verification conditions where needed.
Do not build an unnecessarily complex task management system.
V. Check Before Execution
Before starting development:
- Review the existing outstanding items.
- Check whether the current task involves any of them.
- Determine whether they can be completed along the way.
- If they affect the correctness of the current task, resolve them first.
VI. During Execution
When a new item is discovered:
- If it can be safely completed: complete it directly.
- If it requires additional work: record it.
- If it is temporarily impossible: record the reason.
- If an existing item has been resolved: update its status.
- If an old record turns out to be inaccurate: correct the record.
VII. Check Before Delivery
Before delivery, confirm:
- Are there outstanding items?
- Are there features that are only half completed?
- Are there tests that were not executed?
- Is there documentation that is not in sync?
- Are there problems that were discovered but not yet handled?
- Have they all been recorded?
Partial completion must not be described as full completion.
VIII. Completion Status
Only work that was actually completed may be marked as completed.
A test that did not pass must not be marked as passed.
An operation that was not executed must not be marked as executed.
Functionality that was not verified must not be marked as verified.
IX. Final Principle
Not having told the user does not mean there are no outstanding items.
Unfinished work must have a record and must not disappear silently.
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.

