agentleFS
Sign inStart free
yacb2yacb2/aidex/skills/research/SKILL.md

research

Skill2 starsChanged 33 days agoMITView on GitHub

Use when the user needs to investigate, explore, or spike on how something works and the findings should land as written `.context/research/` notes before any plan or implementation. Fires on "research how X works", "investigate X", "spike on X", "look into how X is done", "I need to understand X before planning". Not for: planning multi-step work once the approach is known (/aidex:plan); recording a decision/ADR (/aidex:decision); capturing a stakeholder request (/aidex:request); documenting a settled, stable reference (/aidex:reference); deferring/parking an idea (/aidex:backlog); ecosystem audits (/aidex:aidex); project-state audits (/aidex:audit).

189 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use daily-arxiv, exp-eval and 10 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.
  • LambenthanSkill

    daily-arxiv

    每日从 arXiv 拉取新论文，过滤相关性，auto-ingest 高优先级论文，检测 SOTA 更新

    Also in 5 projects that use this

  • LambenthanSkill

    exp-eval

    实验判决门：Review LLM 独立评判实验结果 → 4 种判决路径 → 自动更新 claims confidence、ideas status、graph edges

    Also in 5 projects that use this

  • truongtv22Skill

    Problem-Solving Techniques

    Apply systematic problem-solving techniques for complexity spirals (simplification cascades), innovation blocks (collision-zone thinking), recurring patterns (meta-pattern recognition), assumption constraints (inversion exercise), scale uncertainty (scale game), and dispatch when stuck. Techniques derived from Microsoft Amplifier project patterns adapted for immediate application.

    Also in 5 projects that use this

  • liufeng1976Skill

    setup-ts-deep-modules

    Wire dependency-cruiser into a TypeScript repo so each package is a deep module — implementation hidden in subfolders, reachable only through its entry-point files. User-invoked.

    Also in 43 projects that use this

  • skyllwtSkill

    exp-pilot-eval

    Pilot result evaluation — read pilot results, apply success criteria, update idea page (pilot_result, failure_reason if failed), generate PILOT_VERDICT_REPORT. Called by /ideate Phase 5 after /exp-pilot-run.

    Also in 4 projects that use this

  • skyllwtSkill

    exp-pilot-run

    Pilot experiment execution — read Pilot Spec YAML, write pilot code, run experiment(Confirm with the user before operation and require the applicant to conduct manual inspection), return results. Called by /ideate Phase 5. Does NOT modify wiki pages or judge pass/fail.

    Also in 4 projects that use this

  • +5 more from Lambenthan
  • +1 more from skyllwt

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 169 edited versions we read keep these.

  • 0

    Distinctive

    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
  1. Very specific:

    Read the shared documentation canon: ${CLAUDE_PLUGIN_ROOT}/skills/conventions/references/00-global.md (language, IDs, statuses, archive …

    line 18
  2. Very specific:

    If a finding is mechanically verifiable (a benchmark or test the follow-up work should iterate against), link out to an loop loop-spec …

    line 41
  3. Very specific:

    If the project carries a ratchet baseline (.context/.validate-baseline.json), a non-zero exit means you introduced a NEW violation — fix it …

    line 66

“Common” is measured: kept by at least 6 in 10 of the 169 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

Capture an investigation or spike as written notes in .context/research/<topic>/ before committing to a plan or implementation. This skill is the single-purpose entry point for research notes; the shared rules (language, IDs, statuses, archive) live in the conventions canon, but research's own body shape is defined here.

Workflow

  1. Read the shared documentation canon: ${CLAUDE_PLUGIN_ROOT}/skills/conventions/references/00-global.md (language, IDs, statuses, archive rules. Or the .claude/skills/... project-level copy if one exists). Research has its own body shape (below) — it does not borrow the reference runbook/MODULE template.
  2. Pick the shape by size (ADR decision/2026-07-02-research-artifact-shape): a single-document spike is one flat dated file (.context/research/YYYY-MM-DD-<slug>.md — no folder); a multi-document topic is a dated folder (.context/research/YYYY-MM-DD-<slug>/) with 00-index.md (00-overview.md is the accepted alias in research/ only) plus sequential NN-<slug>.md files. When a spike gains a second document, promote it: create the dated folder and move the spike in as 00-index.md.
  3. Front-matter = the global minimum only: title, status, created, updated. No version — a dated spike is a point-in-time investigation, not an evergreen doc that gets revised.

  4. Body = research's own minimal shape, three sections:

    • Question / Scope — what you set out to answer and its bounds.
    • Findings — free-form; number them, subsection them, whatever the investigation needs. No fixed Overview/Prerequisites/Verification skeleton.
    • Answer / Recommendation — mandatory closing section that states what the spike concluded and what it recommends doing next. A spike is not done until this section answers the question.

    If a finding is mechanically verifiable (a benchmark or test the follow-up work should iterate against), link out to an loop loop-spec rather than describing a manual loop here. Write the artifact in English (canon §Language).

Closing a spike

A spike closes when its question is answered — the Answer / Recommendation section is written. On close:

  • Set status: done and stamp updated. done is the terminal state; do not invent answered, complete, or superseded-by-result — those are not in the base status vocabulary the validator enforces.
  • Link what it feeds: escalated_to a plan or decision the research spawned, or set origin_ref on the artifact it produced back to this spike.
  • Research is versioned in place — there is no _archive/ step for research/ (unlike decisions/requests, which archive on close per D-10).

Self-check

Validate the artifact you just wrote and fix any violation before closing:

python3 ${CLAUDE_PLUGIN_ROOT}/skills/conventions/scripts/validate.py --type research

If the project carries a ratchet baseline (.context/.validate-baseline.json), a non-zero exit means you introduced a NEW violation — fix it before closing.

Boundaries

The user wants to…Route to
Plan multi-step / multi-phase implementation workplan
Record a decision / ADRdecision
Capture a stakeholder/client requestrequest
Document a settled system referencereference
Defer / park / shelve an idea for laterbacklog
Audit the Claude Code ecosystemaidex
Audit project state (UX/security/perf/a11y)audit

Related

  • conventions — owns the shared documentation canon (00-global.md: language, IDs, statuses, archive rules).

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.

Check your setup, free