agentleFS
Sign inStart free
YangsonHungYangsonHung/awesome-agent-skills/skills/en/ax-extract-workflow/SKILL.md

ax-extract-workflow

Skill18 starsChanged 6 days agoMITView on GitHub

Use when reconstructing how a shipped feature, commit, date, or artifact was produced from local ax session evidence.

3 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

6 rules. 0 are common, 6 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 6

    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:

    Use this skill when the user asks for:

    line 16
  2. Distinctive:

    Do not use this skill for:

    line 25
  3. Distinctive:

    Commit SHA: use it directly.

    line 34
  4. Distinctive:

    If ax cannot connect to its database, tell the user to start ax services and stop.

    line 56
  5. Distinctive:

    Write the reconstruction from evidence:

    line 57
  6. Distinctive:

    Keep the tone factual. Cite local evidence.

    line 71

“Common” is measured: kept by at least 6 in 10 of the 3 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

ax Extract Workflow

Overview

Reconstruct the workflow behind a shipped artifact from local ax data. Resolve an anchor, find the relevant sessions, inspect turns, tools, skills, commits, and delegations, then write a compact account of what happened.

This skill assumes ax is installed, session data has been ingested, and the local ax database is reachable.

When to Use

Use this skill when the user asks for:

  • How a feature, PR, demo, or artifact was shipped
  • What made a past result work
  • The workflow around a date, commit, topic, or repository event
  • The sessions, tools, skills, or decisions that led to an artifact
  • A repeatable recipe based on prior local agent work

Do not use

Do not use this skill for:

  • Generic recent activity summaries
  • New implementation planning before past evidence is inspected
  • Project status reports not tied to a concrete artifact, date, or commit
  • Reconstructing work when local ax data is unavailable

Instructions

  1. Identify the anchor:
    • Commit SHA: use it directly.
    • Date: use a small window around that date.
    • Topic or artifact name: search commits first.
    • Current repo, recent work: inspect recent sessions for the repo.
  2. Resolve candidate sessions:
ax recall "<topic>" --sources=commit --json
ax sessions near <sha> --json
ax sessions around <date> --days=3 --json
ax sessions here --days=14 --json
  1. Pick up to five sessions. Prefer sessions with high turn counts, matching commits, related files, or relevant tool and skill usage.
  2. Inspect each selected session:
ax sessions show <id> --json
ax sessions show <id> --by-role
ax sessions show <id> --expand=<subagent-uuid>
  1. If ax cannot connect to its database, tell the user to start ax services and stop. Do not invent evidence.
  2. Write the reconstruction from evidence:
    • Ordered workflow arc: what happened first, next, and last.
    • Skill/tool role: what each important skill or tool contributed.
    • Key decisions: user steering points or tradeoffs that changed the work.
    • Repeatable recipe: how to run a similar workflow again.

Output

Return Markdown with:

  • A one-paragraph summary
  • An ordered workflow timeline
  • Key evidence with session ids, commits, or commands
  • A short reproducible recipe when enough evidence exists

Keep the tone factual. Cite local evidence. Mark unknowns instead of filling gaps.

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