agentleFS
urvijain02urvijain02/claude-autopilot/skills/preflight/SKILL.md

preflight

Skill0 starsChanged 24 days agoMITView on GitHub

Pre-run checklist — verify the tools the pipeline needs (gh auth, git, Atlassian MCP, Slack MCP) are connected BEFORE starting. Surfaces anything missing and exactly what will fail or skip, and asks to continue. Used by the autopilot orchestrator at the very start.

  • Installs packages

47 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use adr, ship 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.
  • asterkinSkill

    adr

    Create Architecture Decision Records (ADRs) following project standards. Guides template usage, process, and best practices for documenting architecturally significant decisions.

    Also in 3 projects that use this

  • AWSSkill

    ship

    Finalize dev work on an issue branch — update docs, push, and open a PR.

    Also in 5 projects that use this

  • MicrosoftSkill

    report

    Turn an exploration (threads, findings, charts) into a single Markdown report — note, blog post, executive summary, KPI dashboard, slide brief, or multi-section analytical report, with embedded charts.

    Also in 3 projects that use this

  • Google CloudSkill

    triage

    Gather context on a change or bug and plan the implementation.

    Also in 4 projects that use this

  • AnthropicSkill

    status

    Case status summary by audience — client-facing (plain language), internal (for the professor), or court-ready (formal caption format per local rules). Same facts, different framing and depth. Use when a student needs to update the client, brief the professor, or prepare a court status report.

    Also in 3 projects that use this

  • GitHubSkill

    refactor

    Surgical code refactoring to improve maintainability without changing behavior. Covers extracting functions, renaming variables, breaking down god functions, improving type safety, eliminating code smells, and applying design patterns. Less drastic than repo-rebuilder; use for gradual improvements.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 42 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run this FIRST — before touching the ticket or cloning anything — so missing tools surface up front, not mid-run.

    line 8
  2. Distinctive:

    When a critical CLI tool is missing, surface the recommended 1-click install command for the detected OS (with official URL) inline.

    line 17
  3. Very specific:

    If the user says continue, proceed and skip only the unavailable steps — log each skip in log.md and surface it in the final block (e.g.

    line 53

“Common” is measured: kept by at least 6 in 10 of the 42 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

Preflight

Run this FIRST — before touching the ticket or cloning anything — so missing tools surface up front, not mid-run.

Platform detection & package managers

At run start, probe the host platform:

  • Windows (Git Bash / MSYS / PowerShell): uname -s starts with MINGW*, MSYS*, or CYGWIN* (or $OSTYPE == msys* / PowerShell $env:OS == "Windows_NT").
  • macOS (Darwin): [[ "$(uname -s)" == "Darwin" ]].
  • Linux / WSL: [[ "$(uname -s)" == "Linux" ]] (WSL if /proc/version contains microsoft).

When a critical CLI tool is missing, surface the recommended 1-click install command for the detected OS (with official URL) inline. Do NOT attempt to execute package manager commands automatically (avoids background subshell hangs or UAC/sudo permission stalls).

Check each

NeedHow to checkIf missing → consequence & remediation
platform`uname -s 2>/dev/nullecho "$OS"(matchesMINGW*MSYS*CYGWIN*→ Windows;Darwin→ macOS;Linux` → Linux / WSL)Informs path normalization, package managers, and binary resolution across pipeline stages.
gh authgh auth status 2>&1Critical for remote runs. If missing or unauthenticated: cannot clone private repos or open PRs. Surface recommended OS install command (winget install --id GitHub.cli on Windows, brew install gh on macOS, sudo apt install gh on Linux) or official installer (cli.github.com). (Downgraded to non-blocking warning in offline evaluation / DEMO / --file modes; remote PR step is skipped).
gitcommand -v gitCritical. No clones/branches/commits. Surface recommended OS install command (winget install Git.Git on Windows, brew install git on macOS, sudo apt install git on Linux) or official installer (git-scm.com).
JSON engine. "${CLAUDE_PLUGIN_ROOT}/bin/_paths.sh" && resolve_json_engineIf returns none: Critical. Gate scripts and telemetry cannot parse state JSON. Surface install command for node (winget install OpenJS.NodeJS.LTS), python3, or jq (brew install jq / sudo apt install jq).
Atlassian MCPattempt atlassianUserInfo on whichever Atlassian MCP is connected — the standalone server or the claude.ai connector (mcp__claude_ai_Atlassian__*). Either satisfies this check.Jira steps skip: the ticket won't be assigned or moved.
Slack MCPattempt a lightweight Slack call (e.g. resolve the channel)Slack step skips: the PR link won't be posted.
config${AUTOPILOT_HOME:-~/.claude/autopilot}/config.json exists — check the resolved path, not the literal default, since AUTOPILOT_HOME moves itDefaults are used. A missing slack_channel means the PR-link post is skipped — surface that now rather than at the end of the run.

Report + decide

Print the checklist in this exact shape — one line per check:

preflight <TICKET>
 ✓ platform      windows (git bash)
 ✓ gh auth
 ✓ git
 ✓ engine        jq
 ✗ Atlassian MCP — Jira steps (assign / status transitions) will be skipped
 ✗ Slack MCP     — PR-link post will be skipped
 ✓ config

(✓ <item> when present; ✗ <item> — <consequence> or ✗ <item> — missing; install via: <install-cmd> or <url> when missing.) Then:

  • All present → proceed silently into the run.
  • Anything missing → stop and tell the user what's missing and its consequence:
    • Universal critical miss (git or engine: none) → surface the recommended install command and download link, recommend fixing before running, and do not start (pipeline cannot function without git and JSON parsing).
    • Contextual miss (gh auth):
      • Remote run: do not start; prompt to install or authenticate gh via gh auth login.
      • Offline evaluation / DEMO / --file / --prompt mode: warn that remote PR creation will be skipped, but allow proceeding with local git branches and commits.
    • Optional misses (Atlassian / Slack / config) → state precisely which steps will be skipped, then ask: "Continue anyway?"
  • If the user says continue, proceed and skip only the unavailable steps — log each skip in log.md and surface it in the final block (e.g. jira skipped (Atlassian MCP not connected)). Never skip silently.

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.