write-personal-wiki
Skill4 starsChanged 4 months agoMITView on GitHub
Evaluate and capture durable cross-project knowledge in the personal wiki. Use when work reveals knowledge that may be reusable across repositories, including technical knowledge, general workflows, cross-project gotchas, external research, and non-sensitive internal R&D knowledge.
In this repository14 other files it gives its agents.
- AGENTS.mdAGENTS.md
- build-incrementalSkill
- code-simplifySkill
- create-or-clone-repositorySkill
- design-firstSkill
- implementation-workflowSkill
- manage-git-worktreesSkill
- plan-driven-developmentSkill
- setup-wiki-backed-project-skillsSkill
- subagent-supervisionSkill
- task-briefSkill
- test-driven-debugSkill
- tidy-firstSkill
- write-project-wikiSkill
The file
We split this skill into the rules it gives an agent
12 rules. About 0 are common, 9 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈9
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
Use this skill to evaluate discovered knowledge and document eligible findings in the personal wiki.
- Very specific: line 10
Use ~/agent-wiki/wiki/ unless existing instructions specify otherwise.
- Distinctive: line 17
Remove unnecessary source-project details while preserving essential context.
- Very specific: line 20
Update wiki/index.md when adding an important page.
- Distinctive: line 21
Check links, accuracy, reusability, and safety.
- Distinctive: line 25
Use the existing personal wiki conventions when available.
- Distinctive: line 47
Document knowledge when it is likely to help across repositories or future projects.
- Distinctive: line 74
Keep updates small, factual, self-contained, and easy to review.
- Distinctive: line 76
Prefer updating existing pages over creating duplicates.
- Very specific: line 78
Remove unnecessary project-specific names and context from general categories.
- Distinctive: line 80
Prefer concise summaries over copied text.
- Distinctive: line 82
When uncertain whether a finding belongs in the personal wiki, do not write it automatically and report the question to the user.
≈ “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
Write Personal Wiki
Use this skill to evaluate discovered knowledge and document eligible findings in the personal wiki.
Use ~/agent-wiki/wiki/ unless existing instructions specify otherwise.
Workflow
- Review the knowledge discovered or documented during the task.
- Identify findings that are useful beyond one project.
- Exclude transient, sensitive, or narrowly project-specific information.
- Remove unnecessary source-project details while preserving essential context.
- Integrate the finding into an existing page when possible.
- Otherwise, create a concise page in the appropriate category.
- Update
wiki/index.mdwhen adding an important page. - Check links, accuracy, reusability, and safety.
Personal Wiki
Use the existing personal wiki conventions when available.
The default structure is:
wiki/
index.md
knowledge/
workflows/
gotchas/
internal/Classify content as follows:
knowledge/: Reusable technical or domain knowledge, APIs, SDK usage, data formats, research, implementation patterns, and coding conventions.workflows/: General development, testing, benchmarking, evaluation, deployment, release, and operational procedures.gotchas/: Cross-project pitfalls, external API quirks, environment issues, non-obvious constraints, nondeterminism, and rate limits.internal/: Non-sensitive internal R&D knowledge shared across projects, such as implementation intent, model design, internal interfaces, reusable experimental findings, and relationships between internal systems.
What To Capture
Document knowledge when it is likely to help across repositories or future projects.
Good candidates:
- External research from primary sources, with URL, access date, version, and concise reusable conclusions.
- Reusable API, SDK, data format, configuration, and implementation knowledge.
- General AI/ML workflows involving datasets, training, evaluation, metrics, inference, and artifacts.
- Development workflows and gotchas that apply across projects.
- Non-sensitive internal R&D knowledge that helps transfer implementation intent or findings between projects.
Capture a finding only when it is:
- Useful beyond the current project or task.
- Stable enough to remain useful.
- Safe to retain in the personal wiki.
- Understandable with the context preserved in the page.
Do not document:
- Secrets, credentials, tokens, private data, or sensitive project information.
- Raw logs, temporary errors, or transient task progress.
- One-off project details with no cross-project value.
- Information already documented adequately in an existing page.
- Speculation unsupported by repository evidence or cited sources.
Editing Rules
Keep updates small, factual, self-contained, and easy to review.
Prefer updating existing pages over creating duplicates. Use descriptive kebab-case filenames for new pages.
Remove unnecessary project-specific names and context from general categories. Preserve necessary internal context only under internal/.
Prefer concise summaries over copied text. Link to sources instead of reproducing large amounts of source content.
When uncertain whether a finding belongs in the personal wiki, do not write it automatically and report the question to the user.
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.

