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 skills add gigsmart/haiku-method --skill haiku-reset-stagegit clone --depth 1 https://github.com/gigsmart/haiku-methodWrote 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/skills/gigsmart/haiku-method/haiku-reset-stage)<a href="https://agentmods.dev/skills/gigsmart/haiku-method/haiku-reset-stage"><img src="https://agentmods.dev/badge/skills/gigsmart/haiku-method/haiku-reset-stage/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/gigsmart/haiku-method/haiku-reset-stage"><img src="https://agentmods.dev/badge/skills/gigsmart/haiku-method/haiku-reset-stage.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00057 | $0.00678 |
| Opus 5 | $0.00028 | $0.00339 |
| Sonnet 5 | $0.00011 | $0.00136 |
| Haiku 4.5 | $0.00006 | $0.00068 |
Grade A, and why
haiku-reset-stage 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 9d 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 — 43 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Reset Stage
Wipe ONE stage of an intent and let the engine re-enter it at the elaborate phase on the next tick. The intent's other stages, their approvals, and intent main's history all stay put.
When to use this skill
The intent is fine, but a specific stage produced output that no longer matches what the user wants — usually because the studio's hat instructions, review-agents, or stage config were updated after the stage ran, and the user wants the agent to redo just that stage with the new instructions.
Not the right tool when:
- The intent's whole premise was wrong from the start — that's
/reset-intent. - The stage's output has a few specific problems the user wants flagged — that's a feedback revisit, not a reset.
- The user wants to roll back intent main's git history — that's a manual
git revert/git reset, not in scope for this skill.
How to drive it
When the user says "reset the product stage" (or names one stage):
- Identify the intent. If no intent slug is known from context, call
haiku_intent_listfirst. If multiple are active and the user didn't name one, ask which intent before proceeding — the same discovery preflight/haiku:haiku-reset-intentdoes. - Call
haiku_stage_reset { intent: "<slug>", stage: "<stage-name>" }. The tool confirms via the SPA picker and lists what's about to be deleted. - After the user confirms, the tool performs the wipe (details below).
- The tool returns a message saying to call
haiku_run_next— the next tick re-enters the stage at its elaborate phase. The agent picks up the conversation and re-runs the stage's hat sequence.
What gets wiped
stages/<stage>/units/*.mdstages/<stage>/outputs/,artifacts/,decisions.jsonlstages/<stage>/elaboration.mdstages/<stage>/feedback/*.mdstages/<stage>/discovery/contents (the studio's template files are kept; the agent's produced outputs are wiped)- The stage's git branch (
haiku/<slug>/<stage>). The nexthaiku_run_nextforks it from intent main as needed.
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.
- 9d ago First seen · 43 lines · 57 tokens per session scan A ff332040717b
haiku-reset-stage is a skill published in the GitHub repository gigsmart/haiku-method (24 stars, last pushed 3mo ago), licensed Apache-2.0. It adds 57 tokens to every session and 678 once invoked, about $0.0003 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 skills, from other repositories
contradiction-analysis
A structured way to analyze difficult problems by identifying competing forces, finding the main conflict, and deciding which side or issue has the greatest effect.
mass-line
A method for collecting feedback and facts from multiple people or sources, turning them into a plan, and checking the plan with those affected.
concentrate-forces
A decision-making method for situations where several tasks compete for limited time, attention, computing power, or budget.
practice-cognition
A method for testing ideas in practice, learning from the results, and improving the next attempt through repeated cycles.
protracted-strategy
A long-term planning method for difficult work that cannot be completed quickly, divided into stages with steady progress over time.
redteam-mindset
Red-team operator discipline — the mindset corrections that separate offensive testing from defensive WAPT. Built from authorized red-team work where conservative defaults caused multiple findings to be missed and one to be incorrectly retracted. Use at the START of any red-team engagement and again whenever feeling…