agentleFS
playingpossummplayingpossumm/loose-ends/.claude/skills/bootstrap/SKILL.md

bootstrap

Skill0 starsChanged 33 days agoMITView on GitHub

Interview the user one question at a time to fill mem/ with profile, goals, projects, people, and decision rules. Run once when setting up the vault, and again to refresh. Use when the user says "bootstrap", "set up my brain", "interview me", or when mem/ files still have status "empty".

70 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use surprise-me, ppt-generation and 9 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • stophobiaSkill

    surprise-me

    Create a delightful, unexpected "wow" experience for the user by dynamically discovering and creatively combining other enabled skills. Triggers when the user says "surprise me" or any request expressing a desire for an unexpected creative showcase. Also triggers when the user is bored, wants inspiration, or asks for "something interesting".

    Also in 3 projects that use this

  • HK-hubSkill

    ppt-generation

    Use this skill when the user requests to generate, create, or make presentations (PPT/PPTX). Creates visually rich slides by generating images for each slide and composing them into a PowerPoint file.

    Also in 3 projects that use this

  • stophobiaSkill

    github-deep-research

    Conduct multi-round deep research on any GitHub Repo. Use when users request comprehensive analysis, timeline reconstruction, competitive analysis, or in-depth investigation of GitHub. Produces structured markdown reports with executive summaries, chronological timelines, metrics analysis, and Mermaid diagrams. Triggers on Github repository URL or open source projects.

    Also in 3 projects that use this

  • HK-hubSkill

    consulting-analysis

    Use this skill when the user requests to generate, create, or write professional research reports including but not limited to market analysis, consumer insights, brand analysis, financial analysis, industry research, competitive intelligence, investment due diligence, or any consulting-grade analytical report. This skill operates in two phases — (1) generating a structured analysis framework with chapter skeleton, data query requirements, and analysis logic, and (2) after data collection by other skills, producing the final consulting-grade report with structured narratives, embedded charts, and strategic insights.

    Also in 3 projects that use this

  • MicrosoftSkill

    podcast-generation

    Generate AI-powered podcast-style audio narratives using Azure OpenAI's GPT Realtime Mini model via WebSocket. Use when building text-to-speech features, audio narrative generation, podcast creation from content, or integrating with Azure OpenAI Realtime API for real audio output. Covers full-stack implementation from React frontend to Python FastAPI backend with WebSocket streaming.

    Also in 3 projects that use this

  • VercelSkill

    vercel-deploy

    Deploy applications and websites to Vercel with a claimable preview. Use when the user requests deployment actions like "deploy my app", "deploy and give me the link", "push this live", or "create a preview deployment". No authentication required - returns preview URL and claimable deployment link.

    Also in 4 projects that use this

  • +1 more from HK-hub

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 1 is distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 62 edited versions we read keep these.

  • 1

    Distinctive

    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
  1. Very specific:

    Write the answers to mem/rules.md under a ## Brief heading, then tell them the command that makes it real:

    line 65
  2. Distinctive:

    Never fill a gap with a guess: If they skip something, write that it is unanswered.

    line 79
  3. Very specific:

    Write each mem/*.md as you finish its section, not all at the end, so a stopped interview still leaves something behind.

    line 93

“Common” is measured: kept by at least 6 in 10 of the 62 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

bootstrap

Fill mem/, the self-knowledge store, by interviewing the user. This is the only content in the vault they author rather than you. Everything downstream is generic without it.

Expect roughly twenty minutes. Say so up front, and say they can stop anywhere and resume.

The rule that matters

One question at a time. Wait for the answer before asking the next. Do not batch questions, do not present a form, do not move on until they have actually answered. A dumped list of fifteen questions gets three shallow answers; asked one at a time, the same fifteen get real ones.

Cover, in this order

1. Profile. Who they are, what they do, what the main job is versus what the side work and hobbies are. How they want you to communicate with them. How they work: when they think well, what they avoid, what they are bad at.

2. Goals. For this year. For each: why it matters, what "done" looks like, and roughly when. Push for a checkpoint they could actually miss.

3. Projects. What is active right now. For each: current state, the next concrete action, what is blocking it, and any deadline. This is the file that will go stale fastest.

4. People. Who materially affects their work. Only what they choose to store. Ask before writing anything about a person; store nothing sensitive.

5. Rules. Decision rules, quality standards, boundaries. And explicitly: what must you never do without asking first? That answer becomes the top of rules.md.

6. How the brief should reach them. This one is operational, not biographical, but it decides whether the system ever pushes at all, so ask it rather than assuming.

  • How often? Weekly is the default. Fortnightly suits someone with slow-moving projects; daily almost always becomes noise and gets ignored by week three.
  • When will they read it? Ask this, not when to send it. A brief landing when they have no slack gets archived unread.
  • Then schedule the run for the evening before. This is not a detail, since it decides whether the brief arrives at all.

    A morning task on a laptop that is asleep depends on Windows wake timers, and Windows disables those on battery. The run then waits until the machine is next opened, which may be hours after it was useful. An evening run happens while the machine is already awake and online, and the brief is waiting when they wake up.

    So: reads it Saturday morning → generate Friday 19:00. Reads it Monday → generate Sunday 19:00. Schedule the capture an hour earlier so the brief includes everything from that day.

    The cost is that an evening brief cannot include anything captured overnight. For a periodic review that is not worth the loss of reliability.

  • How should it arrive? Email is set up. If they say "I'll just run it myself", push back once, gently, because a brief you have to remember to run is the exact habit this is meant to replace, and it is the most common way these systems quietly die.
  • How hard should it push? Some people want every open loop every week; others want only what has gone stale. This sets the tone of the whole thing.

Write the answers to mem/rules.md under a ## Brief heading, then tell them the command that makes it real:

python scripts/install_schedule.py --day MON --time 08:00

Substitute their actual answers. Offer to run it for them.

How to ask

  • Push for concrete examples when an answer is vague. "I want to get better at writing" is not usable. "I want to publish one technical post a month, and I know I won't unless something reminds me" is.
  • Never fill a gap with a guess. If they skip something, write that it is unanswered. An honest hole beats an invented fact, and invented facts here poison everything downstream, because this is the file you read every session.
  • Preserve their wording where the phrasing carries meaning. Do not smooth it into business prose.
  • Separate stable from temporary. A role is stable; a current blocker is not. Mark temporary items with stability: temporary and date anything that can expire.
  • Mark inferences as assumptions, never as facts.
  • Do not turn a stated preference into a permanent rule without asking whether it is actually a rule.
  • Refuse to record passwords, keys, credentials, or health data.

Writing the files

Write each mem/*.md as you finish its section, not all at the end, so a stopped interview still leaves something behind. Keep the frontmatter, set status: filled, update updated:.

Append to log.md:

## [YYYY-MM-DD] bootstrap | mem/ filled

Finish with

  1. Which files you wrote and what is in them.
  2. The five most important things you learned.
  3. Five things still missing that would improve the system.
  4. One concrete thing they can do with the brain right now.

Then suggest the obvious first move: capture and ingest their first real source.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.