agentleFS
Baba01hacker666Baba01hacker666/dummyskills/skills/research-brief/SKILL.md

research-brief

Skill0 starsChanged 2 months agoMITView on GitHub

Produces a structured research brief with citations and explicit uncertainty labels from provided notes.

19 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use weekly-review.

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.
  • QuixiAISkill

    weekly-review

    Review the past week across goals, tasks, calendar, journal, and commitments, then propose the next week

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

11 rules. 0 are common, 6 are distinctive, and 5 are very specific to this project.

  • 0

    Common

    Most of the 18 edited versions we read keep these.

  • 6

    Distinctive

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill when:

    line 14
  2. Very specific:

    Do not use to summarize a single document.

    line 20
  3. Very specific:

    Do not use to check isolated claims.

    line 21
  4. Very specific:

    If research_topic is missing: stop and ask: "Please state the research topic or question."

    line 28
  5. Very specific:

    If source_notes is empty: stop and report: "No source notes provided.

    line 29
  6. Distinctive:

    If source notes lack data on a sub-question: explicitly state it under "Known Uncertainties".

    line 30
  7. Distinctive:

    Follow these numbered actions:

    line 33
  8. Distinctive:

    Write a 2-3 sentence overview answering the core question.

    line 35
  9. Very specific:

    If two sources conflict on numbers or behavior: present both numbers with their respective citations and mark confidence as Moderate or Low.

    line 48
  10. Distinctive:

    Never invent a citation index not present in the source notes.

    line 49
  11. Distinctive:

    Stop and ask the user when:

    line 86

“Common” is measured: kept by at least 6 in 10 of the 18 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

Research Brief

Synthesizes multiple investigative notes or search snippets into an executive-level brief. Labels evidence confidence, cites sources explicitly, and identifies unresolved questions.

1. Purpose

Turn raw research findings into an actionable summary. A "Research Brief" answers a specific question, cites findings to source notes, and highlights confidence levels (High, Moderate, Low).

2. When to Use

Use this skill when:

  • Synthesizing findings from multiple documentation pages, articles, or investigation logs.
  • Delivering a research summary to an engineering or product lead.
  • Answering questions that involve incomplete or conflicting data.

3. When Not to Use

  • Do not use to summarize a single document; use factual-summarizer.
  • Do not use to check isolated claims; use claim-checker.

4. Required Inputs

  • research_topic: The core question or topic being investigated (string).
  • source_notes: Numbered or titled notes from investigation (list or text).

5. Missing-Input Behavior

  1. If research_topic is missing: stop and ask: "Please state the research topic or question."
  2. If source_notes is empty: stop and report: "No source notes provided. Please provide investigation notes or links."
  3. If source notes lack data on a sub-question: explicitly state it under "Known Uncertainties".

6. Procedure

Follow these numbered actions:

  1. Action 1: Formulate Executive Summary
    • Write a 2-3 sentence overview answering the core question.
  2. Action 2: Group Findings by Sub-Theme
    • Synthesize evidence with inline citations [Source 1], [Source 2].
  3. Action 3: Assign Confidence Labels
    • High: Verified by multiple sources or authoritative documentation.
    • Moderate: Stated by a single source without conflicting evidence.
    • Low: Inferred, preliminary, or conflicting data.
  4. Action 4: List Known Uncertainties and Recommended Next Steps
    • Highlight open questions and where to look next.
  5. Action 5: Emit Research Brief
    • Format using Section 8.

7. Decision Rules

  • If two sources conflict on numbers or behavior: present both numbers with their respective citations and mark confidence as Moderate or Low.
  • Never invent a citation index not present in the source notes.

8. Output Format

Emit the brief using this template:

### Research Brief: [Topic]

#### Executive Summary
[2-3 sentence synthesis answering the core question.]

#### Key Findings
- **[Finding 1 Title]** (Confidence: [High | Moderate | Low]):
  [Explanation with citations, e.g. [Source 1]]
- **[Finding 2 Title]** (Confidence: [High | Moderate | Low]):
  [Explanation with citations]

#### Known Uncertainties & Gaps
- [Unresolved gap 1]
- [Unresolved gap 2]

#### Recommended Next Steps
1. [Actionable next investigation step]

9. Verification Checklist

  • Every finding has an explicit citation [Source N].
  • Every finding has a confidence rating.
  • Conflicting findings are clearly noted.
  • Uncertainties are explicitly separated from verified facts.

10. Common Failures and Recovery

Failure SymptomCauseRecovery Action
Unsupported assertionsForgetting citationsAppend citation tag [Source N] to every claim.
Artificial certaintyHiding conflicting notesExplicitly document discrepancy under Uncertainties.

11. Stop or Escalation Conditions

Stop and ask the user when:

  • Source notes are contradictory to the point of preventing any coherent summary.

12. Examples

Example 1: Successful Execution

Input:

Topic: Feasibility of HTTP/3 in Node.js
Source 1 (Node docs): Node.js has experimental support for QUIC and HTTP/3 behind a build flag. Not enabled in default binary.
Source 2 (GitHub issue): Community maintains third-party packages, but production stability is unproven for high loads.

Output:

### Research Brief: Feasibility of HTTP/3 in Node.js

#### Executive Summary
HTTP/3 is currently feasible in Node.js only as an experimental preview or via reverse proxy. Core support is not enabled in standard production binaries.

#### Key Findings
- **Core HTTP/3 Support is Experimental** (Confidence: High):
  Node.js core contains QUIC and HTTP/3 support, but it requires building from source with specific flags and is not in standard binaries [Source 1].
- **Production Readiness Concerns** (Confidence: Moderate):
  Third-party modules exist, but issue discussions indicate lack of proven high-load stability in production [Source 2].

#### Known Uncertainties & Gaps
- Exact timeline for official unflagged HTTP/3 support in Node LTS is unstated.

#### Recommended Next Steps
1. Terminate HTTP/3 at edge proxy (Cloudflare, Envoy, or Nginx) rather than Node.js directly.

Example 2: Ambiguous or Missing Input

Input:

Topic: Quantum computing
Source notes: (None)

Output:

### Research Brief: Error - Missing Source Notes

- **Status**: Halted

No source notes were provided for the topic "Quantum computing". Please provide notes, documentation excerpts, or search snippets to synthesize.

Example 3: Expected Final Output

Output:

### Research Brief: SQLite Concurrency Capabilities

#### Executive Summary
SQLite supports concurrent readers with a single active writer when WAL (Write-Ahead Logging) mode is enabled.

#### Key Findings
- **WAL Mode Enables Concurrent Read/Write** (Confidence: High):
  Readers do not block writers and writers do not block readers under WAL mode [Source 1].
- **Single Writer Constraint** (Confidence: High):
  Concurrent write transactions still serialize; multiple simultaneous writers cause busy timeout [Source 1].

#### Known Uncertainties & Gaps
- Maximum transaction throughput under NVMe SSD storage was not benchmarked.

#### Recommended Next Steps
1. Benchmark expected write transactions per second against SQLite busy handler threshold.

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.