Run Experiment
Execute a designed experiment in a worktree and collect structured results. Supports --adjust retry mode.
Install
npx skills add TalonT-Org/AutoSkillit --skill run-experimentgit clone --depth 1 https://github.com/TalonT-Org/AutoSkillit.git /tmp/AutoSkillit
mkdir -p ~/.claude/skills
cp -r /tmp/AutoSkillit/src/autoskillit/skills_extended/run-experiment ~/.claude/skills/run-experimentPersonal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.
From the SKILL.md
# Run Experiment Skill
Execute an experiment that has been implemented in a worktree. This skill
runs whatever the experiment requires — scripts, benchmarks, custom tooling,
manual procedures, data collection, or any combination. It collects results
and produces a structured results file.
The nature of the experiment is entirely determined by the experiment plan.
This skill does NOT prescribe how experiments should be run — it reads the
plan, executes what the plan describes, and reports what happened.
## When to Use
- As the execution step of the `research` recipe (phase 2)
- After `/autoskillit:implement-worktree-no-merge` has set up experiment code
- When `--adjust` flag is passed, re-run with modified approach after a failure
## Arguments
```
/autoskillit:run-experiment {worktree_path} [--adjust]
```
- `{worktree_path}` — Absolute path to the worktree containing experiment code
(required). Scan tokens for the first path-like token (starts with `/`, `./`,
or `.autoskillit/`).
- `--adjust` — Optional flag indicating this is a retry after a previous failure.
When present, read the previous results/errors from `{{AUTOSKILLIT_TEMP}}/run-experiment/`
and adjust the approach before re-running.
## Critical Constraints
**NEVER:**
- Modify files outside the worktree
- Merge the worktree — leave it intact for the orchestrator
- Skip result collection — every run must produce structured output
- Assume what kind of experiment this is — read the plan and follow itDetails
- Skill name
run-experiment- Repository
- TalonT-Org/AutoSkillit
- Path
src/autoskillit/skills_extended/run-experiment/SKILL.md- License
- MIT
- Repo updated
- 8 November 2026
FAQ
What does the Run Experiment skill do?
Execute a designed experiment in a worktree and collect structured results. Supports --adjust retry mode.
How do I install the Run Experiment skill?
skills CLI (Claude Code, Cursor, Codex and more): npx skills add TalonT-Org/AutoSkillit --skill run-experiment. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.
Is Run Experiment popular?
It lives in TalonT-Org/AutoSkillit, which has 5 GitHub stars and 1 forks.
Related skills
Analyze Prs
5TalonT-Org/AutoSkillit
Analyze all open PRs targeting a base branch — determine merge order, identify file overlaps, and tag each PR as simple or needs_check for complexity. Use at the start of a PR consolidation workflow.
Arch Lens C4 Container
5TalonT-Org/AutoSkillit
Create C4 Container architecture diagram showing static structure, building blocks, and technology choices. Anatomical lens answering "How is it built?
Arch Lens Concurrency
5TalonT-Org/AutoSkillit
Create Concurrency architecture diagram showing parallel execution patterns, thread pools, synchronization, and barriers. Physiological lens answering "How does parallelism work?
Triage Issues
5TalonT-Org/AutoSkillit
Analyze open GitHub issues and produce a sequenced implementation plan — grouping issues into parallel batches, ordering those batches, and tagging each issue with its recipe route. Use when user says "triage issues", "prioritize issues", or "plan issue order".
Review Research PR
5TalonT-Org/AutoSkillit
Automated diff-scoped research PR review using parallel audit subagents aligned to research quality dimensions. Posts inline GitHub review comments and submits a summary verdict. Use after a research PR is opened to gate on review approval.
Contributor Experience Review
1priyank766/OpenSource-SKILL
Review of a first-time or early-stage contributor's pull request, where the technical bar stays fixed but the sequencing, batching, and supplied context change. Use when GitHub labels the author a first-time contributor, when the PR comes from a newcomer to this repo, when it arrives via Hacktoberfest, GSoC, Outreachy, or a good-first-issue, when the change is technically close but procedurally messy (missing DCO, wrong base branch, unsigned commits, failing lint), or when the PR must be declined and the decline needs to not cost the project a future contributor. Applies a raised severity floo
Browse the most popular Claude skills, all git & code review skills, or write your own with the skills generator.