Skip to content
Agentmart
Submit a tool

discovery

Frame an uncertain product problem into evidence-aware problem hypotheses, learning questions, validation approaches, and decision criteria.

Skill by ThorStarlord

IndexedMIT
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/thorstarlord-discovery

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Frame an uncertain product problem into evidence-aware problem hypotheses, learning questions, validation approaches, and decision criteria. Use when the active agent has a problem-discovery responsibility, the problem is vague or contested, or investment should be preceded by bounded learning. Produce discovery_findings without treating a research plan as proof that a hypothesis was validated.

Use cases: research, agent-building

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/discovery

What's inside

Skill name
discovery
Description
Frame an uncertain product problem into evidence-aware problem hypotheses, learning questions, validation approaches, and decision criteria. Use when the active agent has a problem-discovery responsibility, the problem is vague or contested, or investment should be preceded by bounded learning. Produce discovery_findings without treating a research plan as proof that a hypothesis was validated.
License
MIT
Instructions
38 lines in SKILL.md · view on GitHub

Files (3)

  • SKILL.md
  • methodology.md · references/
  • output-contract.md · references/

SKILL.md (first lines)

# Discovery

Satisfy a `problem_discovery` responsibility by producing `discovery_findings`.

## Inputs

Require a problem statement, relevant product/customer context, and known constraints. Consume supplied persona, customer, repository, market, operational, or experiment evidence when available.

## Procedure

1. Read [references/methodology.md](references/methodology.md).
2. Reframe the problem in user/job/outcome terms without smuggling a solution into the problem statement.
3. Separate established observations from candidate root-cause hypotheses.
4. Identify decision-relevant hypotheses; use 3-5 only when that breadth is warranted.
5. For each hypothesis, identify evidence already available, missing evidence, the decision that evidence could change, and the lowest-cost plausible evidence source. A test/observation is only one possible source.
6. When empirical evidence is material, apply `../using-sensemaking/references/experiment-economy-v1.md` before recommending a probe or experiment. Do not promote missing evidence directly into experimentation.
7. Sequence learning by decision-relevant uncertainty reduction and total evidence-acquisition cost, not by presentation convenience.
8. Render [references/output-contract.md](references/output-contract.md) and return control.

## Stop or downgrade

- A proposed experiment is not validation evidence.
- If the requested conclusion requires customer or operational evidence that is unavailable, leave the hypothesis unresolved and name the missing evidence.
- Do not launch experiments, contact users, or mutate external systems without separate authority.

## Boundary

```text
uncertainty != experiment
missing evidence != experiment required
discovery recommendation != experimentation responsibility
validation plan != validated hypothesis
well-structured discovery artifact != correct product direction
```

Discovery owns the evidence need and candidate evidence source. Experiment Economy owns whether experimentation is warranted.

From ThorStarlord/sensemaking-skills, MIT.

More from ThorStarlord/sensemaking-skills

Signals

Stars
0
Forks
0
Open issues
7
Last commit (this folder)
Sep 21, 2026 (1 month ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from ThorStarlord

Cite this page

discovery skill. Agent Mart. https://agentmart.dev/skills/thorstarlord-discovery (last updated 2026-10-24)