Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/minhoyoo-iotrust/claude-plugin/autopilotgit clone --depth 1 https://github.com/minhoyoo-iotrust/claude-pluginWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/minhoyoo-iotrust/claude-plugin/autopilot)<a href="https://agentmods.dev/commands/minhoyoo-iotrust/claude-plugin/autopilot"><img src="https://agentmods.dev/badge/commands/minhoyoo-iotrust/claude-plugin/autopilot.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00017 | $0.03482 |
| Opus 5 | $0.00009 | $0.01741 |
| Sonnet 5 | $0.00003 | $0.00696 |
| Haiku 4.5 | $0.00002 | $0.00348 |
Grade A, and why
autopilot scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 453 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CRITICAL INSTRUCTION: Do NOT use the Skill tool to invoke gsdu:autopilot. Do NOT re-invoke this command. All workflow steps are below — execute them directly.
For each unfinished phase: plan (if needed) → execute → verify → gap closure (if needed). After all phases complete: audit milestone → handle gaps → report.
Orchestrator stays lean (~15% context). Each phase plan/execute is delegated to subagents with fresh context.
Prerequisites:
- GSD must be installed (
~/.claude/get-shit-done/exists) - A milestone roadmap must exist (detected via
gsd-tools.cjs init progress— supports both standard and workstream layouts) - At least one phase must be unfinished
# 1. Check GSD installation
test -f "$HOME/.claude/get-shit-done/bin/gsd-tools.cjs" || echo "GSD_NOT_FOUND"
If GSD_NOT_FOUND:
GSD is not installed. Install it first:
https://github.com/gsd-build/get-shit-done
Then run /gsdu:autopilot again.
Exit.
# 2. Load context and detect paths (supports both standard and workstream layouts)
INIT=$(node $HOME/.claude/get-shit-done/bin/gsd-tools.cjs init progress)
Extract path variables from INIT JSON — these adapt automatically to workstream mode:
ROADMAP_PATH←roadmap_path(e.g.,.planning/ROADMAP.mdor.planning/workstreams/{name}/ROADMAP.md)STATE_PATH←state_path(e.g.,.planning/STATE.mdor.planning/workstreams/{name}/STATE.md)CONFIG_PATH←config_path(e.g.,.planning/config.jsonor.planning/workstreams/{name}/config.json)PLANNING_DIR← parent directory ofROADMAP_PATH(e.g.,.planning/or.planning/workstreams/{name}/)PHASES_DIR←{PLANNING_DIR}/phases/ROADMAP_EXISTS←roadmap_exists
If ROADMAP_EXISTS is false:
No milestone roadmap found.
Start a new milestone first:
/gsd:new-project — for new projects
/gsd:new-milestone — for existing projects
Exit.
# 3. Analyze roadmap for phase details
ROADMAP=$(node $HOME/.claude/get-shit-done/bin/gsd-tools.cjs roadmap analyze)
Extract from INIT and ROADMAP JSON:
phases[]— all phases withnumber,name,goal,disk_status,plan_count,summary_countmilestone_versioncompleted_count,phase_count
Build execution queue:
- Filter phases where
disk_statusis NOTcomplete - If
start-phaseargument given: skip phases before that number - Sort by phase number ascending
If queue is empty:
All phases are complete! Run milestone audit:
/gsd:audit-milestone
Exit.
## Autopilot Execution Plan
**Milestone:** v{milestone_version}
**Progress:** {completed_count}/{phase_count} phases complete
### Phases to Execute
| # | Phase | Status | Plans | Action |
|---|-------|--------|-------|--------|
| {N} | {name} | {disk_status} | {plan_count} | {plan if 0, execute if >0} |
| ... | ... | ... | ... | ... |
### After All Phases
1. Milestone audit
2. Gap closure (if needed, max {MAX_GAP_ITERATIONS} iterations)
### Safety Rails
- Phase gap closure: max {MAX_GAP_ITERATIONS} iterations per phase
- Milestone gap closure: max {MAX_GAP_ITERATIONS} iterations
- checkpoint:human-action always pauses
- Errors stop execution with progress report
Run without --dry-run to execute.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago First seen · 453 lines · 17 tokens per session scan A d395d4d29976
autopilot is a command published in the GitHub repository minhoyoo-iotrust/claude-plugin (5 stars, last pushed 4mo ago), licensed MIT. It adds 17 tokens to every session and 3,482 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.