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/simonrowland/goal-flight/doctorgit clone --depth 1 https://github.com/simonrowland/goal-flightWrote 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/simonrowland/goal-flight/doctor)<a href="https://agentmods.dev/commands/simonrowland/goal-flight/doctor"><img src="https://agentmods.dev/badge/commands/simonrowland/goal-flight/doctor.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 5.1 | $0.00018 | $0.01698 |
| Opus 5 | $0.00009 | $0.00849 |
| Sonnet 5 | $0.00004 | $0.00340 |
| Haiku 4.5 | $0.00002 | $0.00170 |
Grade A, and why
doctor scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 137 lines — stays where its author put it; the contents beside it link to each section on GitHub.
doctor
Diagnose whether goal-flight can run safely in the current session. Read
protocols/tool-readiness.md, protocols/project-state-layout.md,
protocols/task-lifecycle.md, and protocols/progress-dashboard.md, then run
the procedural doctor.
Command
python3 <skill-root>/scripts/goalflight_doctor.py --project-root "$PWD"
For machine-readable output:
python3 <skill-root>/scripts/goalflight_doctor.py --project-root "$PWD" --json
For a live worker write-file probe:
python3 <skill-root>/scripts/goalflight_doctor.py --project-root "$PWD" --worker-write-probe --write-probe-agent grok-code
Required Checks
The script owns the checks; this file names the contract so tests and humans can verify drift. Doctor separates host-global install readiness from project-local Goal Flight readiness.
Host-global checks cover PATH/binary presence, selected CLI versions, host-wrapper installation, context-mode registration, gstack, autoreview, capacity, model currency probes, and rate-pressure signals.
Project-local checks cover docs-private/env-caveats.md, the canonical
docs-private/ state tree, repository SKILL.md, AGENTS.md Goal Flight
routing, the AGENTS.md newest-RESUME-NOTES living-state pin, project
verification commands, resume notes, missing state files, managed-view
schema/template skew, and the tasks.jsonl <-> tasks-data.js mirror.
Package plugin validation applies only when --project-root is the Goal Flight
package repository; for normal target projects it is skipped as INFO.
- Validate goal-flight packaging and adapter manifests. Host-native validators are adapter-owned compatibility probes, not the core readiness source.
- Check Codex Desktop:
/Applications/Codex.appmdfind 'kMDItemCFBundleIdentifier == "com.openai.codex"'
- Check
codex --version. - If Codex Desktop exists but
codexCLI is missing, WARN with:Codex Desktop found, but codex CLI missing. Install CLI with npm install -g @openai/codex && codex login. Desktop install implies the user likely already has an OpenAI account; CLI login should use that account. - Check Codex context-mode registration with
scripts/register-context-mode-codex.py --check. - Check Cursor context-mode registration with
scripts/register-context-mode-cursor.py --scope global --checkandscripts/register-context-mode-cursor.py --scope project --project-root "$PWD" --check. - Check
gstack --version. - Check autoreview:
scripts/autoreview.shexecutable and vendored helper atautoreview/scripts/autoreview(AUTOREVIEW_HELPERoverrides). - Check first-class local worker/controller candidates:
- Codex: Desktop/CLI present, context-mode registered when large-output work will run, ACP adapter binary present when the Codex ACP path is considered.
- Cursor: Desktop,
cursor, andcursor-agentpresent when Cursor is routed as orchestrator or worker; global or projectmcp.jsonhas context-mode when large-output work will run./Applications/Cursor.appcommand -v cursorcommand -v cursor-agent~/.cursor/mcp.json.cursor/mcp.json
- Grok: Grok Build present, headless flags available. File-writing routes
require the live write-file e2e probe to pass in the current environment;
otherwise use Grok for read-only analysis/research only.
--prompt-file--permission-mode--os-sandbox- Each configured seat home must have
permission_modein that home's.grok/config.toml. Doctor warns and dispatch refuses when the key is absent; do not add a grok CLI--permission-modeflag to compensate.
- Claude compatibility path: CLI/plugin checks pass before Claude-specific
compatibility examples are used. NOTE: presence/PATH only — a green doctor
does NOT confirm the remote claude-acp credential. Headless subscription
seat readiness is verified per node by the fleet auth probe
(
claude auth status --json), not by host-global doctor; pty headroom is covered bypty_shim_health(orphan reaping), not a pre-dispatch free-pty gate.
- Check ACP worker adapters for presence/PATH only:
codex-acp,cursor-agent,claude-code-cli-acp,grok agent stdio. - Check
claude_acp_stopgap; when the stopgap is required but absent, usescripts/install_claude_acp_patch.shand re-run doctor. - Check project git state.
- Check project Goal Flight readiness:
docs-private/env-caveats.md- canonical
docs-private/files/dirs fromtemplates/state-skeleton/andprotocols/project-state-layout.md - root
SKILL.md AGENTS.mdGoal Flight routing with skill-root/load-order guidanceAGENTS.mdpins the newestdocs-private/RESUME-NOTES-*.mdliving state- project test/lint/build commands
- optional
docs-private/RESUME-NOTES*.md - managed static view assets match the templates from
templates/state-skeleton/; schema/template skew is a layout warning docs-private/tasks.jsonlanddashboard/tasks-data.jsmirror validation when either task-store file exists
- Check machine capacity profile.
- Cursor model currency (
cursor_models_probe): runscursor-agent models, identifies the leading internalcomposer-X.Y(non--fast), compares against~/.cursor/cli-config.jsonmodelId. Flagsuser_behindwhen the user is on an older internal model or on a paid-passthrough model. Surfaces as[OK]or[WARN]with the cli-config edit recommendation. - Worker CLI currency (
worker_currency_probe): proxy for model currency for workers without a native list-models API.grok update --check --jsonfor grok;npm view <pkg> versionfor@openai/codex/@anthropic-ai/claude-code/claude-code-cli-acp. Behind workers get[WARN]; verified-current workers[OK]; probe-failed (registry unreachable) workers[INFO]as a separate line. - Rate-pressure summary (
_rate_pressure_summary→goalflight_rate_pressure): scans the dispatch ledger for provider-level rate-limit signatures over the last 10 minutes.[WARN]per pressured provider with recommended caps and fallback providers;[OK]when all providers clear. - PTY shim health (
pty_shim_health): orphaned claude-acp shims and pty cap pressure. Remediation:bin/gf-reap-shims --exec(orpython3 scripts/goalflight_reap_shims.py --exec).
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.
- today Changed · +10 tokens per session ea1b0b38185a
- 6d ago First seen · 137 lines · 8 tokens per session scan A e945217fb399
doctor is a command published in the GitHub repository simonrowland/goal-flight (20 stars, last pushed yesterday), licensed MIT. It adds 18 tokens to every session and 1,698 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 2026-08-30.
Other commands, from other repositories
review-context7-updates
Please read the @README.md and @docs/tools/package-documentation.md to understand the high level context of my repo.
mine
Quickly capture relevant project context into Memento memory.
help
Show comprehensive Memento help — available MCP tools and usage patterns.
init
Set up Memento — configure MCP server and verify everything works.
search
Search your memories using Memento hybrid retrieval.
status
Show Memento server status and current configuration.