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/davidmatousek/tachi/aod.plangit clone --depth 1 https://github.com/davidmatousek/tachiWrote 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/davidmatousek/tachi/aod.plan)<a href="https://agentmods.dev/commands/davidmatousek/tachi/aod.plan"><img src="https://agentmods.dev/badge/commands/davidmatousek/tachi/aod.plan.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 | $0.00021 | $0.00562 |
| Opus 5 | $0.00010 | $0.00281 |
| Sonnet 5 | $0.00004 | $0.00112 |
| Haiku 4.5 | $0.00002 | $0.00056 |
Grade A, and why
aod.plan 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.
Copies of this mod
1 near-identical copy found in the catalogue:
- aod.plan — 100% identical, 0 lines differ
How it starts
The opening of the file, as written. The whole thing — 45 lines — stays where its author put it; the contents beside it link to each section on GitHub.
User Input
$ARGUMENTS
Consider user input before proceeding (if not empty).
Overview
Orchestrates all three Plan sub-steps in sequence. Advances automatically when governance reviews pass (APPROVED). Stops if any review returns CHANGES_REQUESTED or BLOCKED, so the user can fix and re-run /aod.plan to resume from where it stopped.
Flow: Detect branch → Read artifact states → Run sub-steps in sequence (spec → project-plan → tasks) → Stop on rejection or report completion
Execution
Invoke the ~aod-plan skill to perform orchestration. Pass any user arguments through to the sub-commands.
The skill will:
- Determine the current feature from the git branch
- Check spec.md, plan.md, and tasks.md for existence and approval status
- Start from the first sub-step that needs work (skips already-approved artifacts)
- After each sub-step is approved, automatically advance to the next
- Stop on CHANGES_REQUESTED or BLOCKED — report what needs fixing
- Report "Plan stage complete" when all three artifacts are approved
Common Rationalizations
| Rationalization | Reality |
|---|---|
"I'll bypass /aod.plan and run the sub-commands myself" |
/aod.plan resumes from the first unfinished sub-step (Step 3 of this file) and skips already-approved artifacts. Running sub-commands manually loses auto-advance and the artifact-state check. |
"The spec was rejected — I'll re-run /aod.plan to skip past it" |
Step 5 stops on CHANGES_REQUESTED by design and re-enters at the rejected sub-step. Address review notes; the orchestrator does not bypass rejections. |
"I'll run /aod.plan from main — it will figure out the branch" |
Step 1 derives the feature from git branch --show-current. From main, no NNN-* feature is detected and the orchestrator aborts. Create the feature branch first. |
Red Flags
- Agent runs
/aod.spec,/aod.project-plan, and/aod.tasksseparately without invoking/aod.planonce. - Agent re-invokes
/aod.planimmediately after a CHANGES_REQUESTED stop without editing the rejected artifact. - Agent invokes
/aod.planfrom themainbranch instead of aNNN-*feature branch. - Agent reports "Plan stage complete" without the Step 6 confirmation message from the orchestrator.
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 · 45 lines · 21 tokens per session scan A 28c908b94df0
aod.plan is a command published in the GitHub repository davidmatousek/tachi (90 stars, last pushed 22d ago), licensed Apache-2.0. It adds 21 tokens to every session and 562 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
fortigate
Run cs-security-analyst (AT + CA workflows) against the Fintech FortiGate zero-day scenario and produce a 7-task scorecard.
README
Claude Code slash commands for the Unified Security Agent Platform (USAP). These commands let you load any USAP skill as a live LLM persona and run structured compliance tests — all inside a Claude Code session.
challenge
Run cs-security-analyst (AT workflow) against the Perfect Storm 8-vector crisis scenario and produce a 12-check mock comparison scorecard.
compare
Before/after comparison of zero-day-response SKILL.md v1 (broken) vs v2 (fixed) against the FortiGate zero-day scenario. Outputs a scored table.
run
Load a USAP skill SKILL.md and activate it as your operating persona. Argument: skill slug.
test
Run a USAP skill against the FortiGate zero-day test scenario and produce a 6-problem compliance scorecard.