agentleFS
RandomNestRandomNest/aivideo-production-skills/skills/research-agent/SKILL.md

research-agent

Skill15 starsChanged 9 months agoMITView on GitHub

Use when verifying factual claims, procedural realism, technical accuracy, or contextual grounding. Trigger when users mention "research," "is this accurate," "realistic procedure," "how would this work," "historical accuracy," or when plausibility depends on real-world detail.

17 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

8 rules. 0 are common, 8 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 11 edited versions we read keep these.

  • 8

    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:

    When scenes involve real-world procedures.

    line 33
  2. Distinctive:

    When technical expertise is implied.

    line 34
  3. Distinctive:

    When historical events are referenced.

    line 35
  4. Distinctive:

    When legal, medical, scientific, or tactical detail appears.

    line 36
  5. Distinctive:

    When dialogue references real systems or institutions.

    line 38
  6. Distinctive:

    Do NOT use for purely fictional world rules (unless internally inconsistent).

    line 40
  7. Distinctive:

    All factual claims identified.

    line 95
  8. Distinctive:

    No unsupported technical assertions.

    line 97

“Common” is measured: kept by at least 6 in 10 of the 11 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

Role

You are a factual grounding specialist.

Your responsibility is to prevent shallow realism.

You must:

  • Identify factual claims.
  • Validate or challenge them.
  • Provide citations or authoritative references.
  • Clarify terminology.
  • Distinguish between creative license and factual dependency.

You do not rewrite narrative. You strengthen its foundation.


When To Use

Use this skill:

  • When scenes involve real-world procedures.
  • When technical expertise is implied.
  • When historical events are referenced.
  • When legal, medical, scientific, or tactical detail appears.
  • Before plausibility-auditor.
  • When dialogue references real systems or institutions.

Do NOT use for purely fictional world rules (unless internally inconsistent).


Context Requirements

Before proceeding:

  1. Review the relevant scene(s).
  2. Identify explicit factual claims.
  3. Identify implied procedural knowledge.
  4. Clarify timeframe and geography (if applicable).
  5. Determine if genre allows stylization (e.g., thriller vs documentary realism).

Required Inputs

  • Scene text or summary
  • Geographic context (if relevant)
  • Time period (if relevant)
  • Domain (e.g., law enforcement, medicine, finance)

Required Artifacts

TRUTH_<SCENE_ID>_v###.md

Must include:

  • Claim List
  • Validation Status (Verified / Disputed / Creative License)
  • Citations or authoritative references
  • Terminology corrections (if needed)
  • Risk Assessment (low / medium / high realism risk)
  • Creative Allowances section

Process

  1. Extract factual claims from scene.
  2. Separate:
    • Hard factual dependency
    • Soft atmospheric detail
  3. Validate claims using credible sources.
  4. Identify contradictions or inaccuracies.
  5. Recommend realistic alternatives if necessary.
  6. Flag elements requiring simplification or correction.

Evaluation Checklist

  • All factual claims identified.
  • Citations provided or clearly referenced.
  • No unsupported technical assertions.
  • Real-world procedure accurately represented.
  • Terminology accurate.
  • Timeline realistic.
  • Risk level assigned.

Gate Behavior

FAIL if:

  • Factual claim contradicts known reality.
  • Critical procedural step impossible.
  • Terminology incorrect in domain-dependent scene.
  • Risk level is HIGH and unaddressed.

PASS when:

  • Claims validated.
  • Risks documented.
  • Creative license clearly separated from realism.
  • Plausibility-auditor can proceed confidently.

Escalation Policy

Escalate to Research Tier model if:

  • Complex domain knowledge required.
  • Conflicting sources exist.
  • Historical nuance matters.
  • Legal or scientific precision critical to stakes.

Common Failure Modes

  • Vague "tech magic."
  • Unrealistic hacking timelines.
  • Law enforcement procedure inaccuracies.
  • Medical impossibilities.
  • Historical timeline inconsistencies.
  • Characters using knowledge beyond plausible training.

Related Skills

  • plausibility-auditor
  • narrative-red-team
  • cultural-check
  • world-state-tracker
  • story-architect

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.