agentleFS
malinskibeniaminmalinskibeniamin/skills/brain-dump/SKILL.md

brain-dump

Skill4 starsChanged 10 days agoMITView on GitHub

Turn raw thoughts, monologues, notes, files, or links into a multi-opportunity brief before grilling.

8 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 7 edited versions we read keep these.

  • 1

    Distinctive

    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
  1. Distinctive:

    Keep it self-contained; link or cite sources instead of copying them.

    line 64

“Common” is measured: kept by at least 6 in 10 of the 7 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

Optional discovery before /grilling. Unstructured thoughts or a monologue may reveal several independent tracks; preserve them instead of forcing one objective.

Return the brief in chat unless asked to save or publish. Do not implement, create tickets, or commit to opportunities.

1. Absorb

Treat the conversation, notes, attachments, articles, files, and links as one dump. If the user is still talking, acknowledge and wait; otherwise proceed. Read supplied material and repo evidence. Extract external claims, conclusions, constraints, timing, and implications; use /read-the-damn-docs or /research when current primary sources matter.

Separate source/repo facts, user observations, labelled inferences, contradictions, and unknowns. Never ask the user to repeat an answer already present.

2. Reconstruct

Extract actors, pains, outcomes, systems, triggers, constraints, ideas, rejected directions, urgency, and success signals. Name the surface; separate need from solution. Record implied answers in an Answer ledger:

  • Settled: explicit or directly supported.
  • Tentative: inferred and safe to challenge.
  • Unknown: missing and able to change direction.

3. Map opportunities

Cluster duplicates and list each distinct, supported direction. Usually show 2-5; keep more only for genuinely broader work. Include disciplines only when evidence supports them.

For each opportunity give outcome/actor, evidence, work products, dependencies/risks/decisions, and cheapest proof: lookup, prototype, measurement, or reversible slice. Recommend by value, evidence, urgency, and reversibility; keep alternatives visible.

4. Return

Prune empty sections from this structure:

## Brain dump brief

### Orientation
<surface, tension, recommendation>

### Source synthesis
<facts, implications, contradictions, citations or paths>

## Answer ledger
| Likely grilling question | Extracted answer | State | Evidence |
|---|---|---|---|
| ... | ... | Settled / Tentative / Unknown | ... |

## Opportunity map
### <Opportunity>
- Outcome:
- Evidence:
- Work products:
- Risks and dependencies:
- Cheapest next proof:

## Grilling handoff
- Settled context to preserve:
- Tentative assumptions to challenge:
- Material user decisions still open:
- Facts to look up without asking the user:
- Candidate prototypes or measurements:

Keep it self-contained; link or cite sources instead of copying them.

5. Hand off

Continue with /grilling for material user decisions and pass every track. Ask only Unknown items that could invalidate or prioritize it; challenge Tentative items when downside matters; preserve Settled answers unless contradicted. Otherwise recommend the next lookup, prototype, spec, plan, or execution skill.

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.