agentleFS
groupzer0groupzer0/vs-code-agents/.github/skills/analysis-methodology/SKILL.md

analysis-methodology

Skill291 starsChanged 10 months agoMITView on GitHub

A systematic approach to converting unknowns to knowns through structured investigation. Objective: Every analysis session should reduce uncertainty. If an unknown cannot be resolved, document why and what is needed to close it. Classify every finding by its evidence strength: Rule: Findings at Level 3 (Inferred) should be flagged for upgrade to Level 1 (Proven) before being used for decisions. Use this structure to surface remaining unknowns: Behaviors: - Populate this table during investigation, not just at the end. -…

1 other project uses a version of this, 1 of them edited.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 1

    1. gopalparashar421/agent-skills★ 243% same

    Wrote their own version · 0

    None found.

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 9 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈9

    Distinctive · estimate

    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:

    Objective: Every analysis session should reduce uncertainty.

    line 7
  2. Distinctive:

    Use this structure to surface remaining unknowns:

    line 27
  3. Distinctive:

    Use these patterns to move unknowns to knowns:

    line 47
  4. Distinctive:

    Add targeted logging to isolate behavior.

    line 50
  5. Distinctive:

    Use mocks/stubs to eliminate variables.

    line 55
  6. Distinctive:

    Write minimal, runnable code to prove or disprove a behavior.

    line 62
  7. Distinctive:

    Follow data/control flow backward to find the root cause.

    line 66
  8. Distinctive:

    Ask: "Where did this value *come from*?"

    line 67
  9. Distinctive:

    Before handing off to another agent or user:

    line 88

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “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

Analysis Methodology

A systematic approach to converting unknowns to knowns through structured investigation.

Core Principle

Objective: Every analysis session should reduce uncertainty. If an unknown cannot be resolved, document why and what is needed to close it.


Confidence Levels

Classify every finding by its evidence strength:

LevelLabelMeaningExample
1ProvenVerified by code execution, POC, or reproducible test."Running npm test -- --filter=X confirms the error."
2ObservedSeen in logs, monitoring, or direct inspection, but not isolated."The error appears in production logs at 3am."
3InferredDerived from documentation, patterns, or logical deduction."The API docs suggest this should return 404."

Rule: Findings at Level 3 (Inferred) should be flagged for upgrade to Level 1 (Proven) before being used for decisions.


Gap Tracking Template

Use this structure to surface remaining unknowns:

## Remaining Gaps

| # | Unknown | Blocker | Required Action | Owner |
|---|---------|---------|-----------------|-------|
| 1 | Why does X fail under load? | Cannot reproduce locally. | Need staging access or load test harness. | [TBD] |
| 2 | Does API Y support pagination? | Docs unclear. | Contact vendor or run POC against sandbox. | [TBD] |

Behaviors:

  • Populate this table during investigation, not just at the end.
  • Surface the table to the user in chat before handoff.
  • Mark items as "Resolved" (with link to finding) or "Deferred" (with rationale).

Investigation Techniques

Use these patterns to move unknowns to knowns:

Log Tracing

  • Add targeted logging to isolate behavior.
  • Compare expected vs actual log sequences.

Component Isolation

  • Reproduce behavior with minimal dependencies.
  • Use mocks/stubs to eliminate variables.

Binary Search Debugging

  • Narrow the failure window by bisecting commits, config changes, or code paths.
  • Halve the search space with each step.

POC Execution

  • Write minimal, runnable code to prove or disprove a behavior.
  • POCs must be reproducible by others (check in or share).

Upstream Tracing

  • Follow data/control flow backward to find the root cause.
  • Ask: "Where did this value come from?"

Analysis Document Structure

Recommended sections for analysis outputs:

  1. Changelog — Date, handoff context, outcome summary.
  2. Value Statement & Objective — Why this analysis matters.
  3. Context — Background, scope, constraints.
  4. Methodology — What techniques were used (reference above).
  5. Findings — Factual results, classified by Confidence Level.
  6. Gap Tracking Table — (see template above).
  7. Analysis Recommendations — Next steps to deepen inquiry (not solutions).
  8. Open Questions — Unresolved items requiring user/agent input.

Handoff Protocol

Before handing off to another agent or user:

  1. List resolved unknowns — What was determined and at what confidence level.
  2. List remaining gaps — Use the Gap Tracking Table.
  3. State blockers explicitly — What prevents further progress.
  4. Communicate in chat — Do not rely solely on the document; surface gaps directly.

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.