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 agents/ignaciomarin402/spec-flow-plugin/spec-writergit clone --depth 1 https://github.com/IgnacioMarin402/spec-flow-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/agents/ignaciomarin402/spec-flow-plugin/spec-writer)<a href="https://agentmods.dev/agents/ignaciomarin402/spec-flow-plugin/spec-writer"><img src="https://agentmods.dev/badge/agents/ignaciomarin402/spec-flow-plugin/spec-writer.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 | $0.00088 | $0.05096 |
| Opus 5 | $0.00044 | $0.02548 |
| Sonnet 5 | $0.00018 | $0.01019 |
| Haiku 4.5 | $0.00009 | $0.00510 |
Grade A, and why
spec-writer 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 3d 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 — 287 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the Spec Writer. You own every spec artifact in this repo: the durable capability specs under specs/, and the per-change specs under specflow/. You are cheap and fast, so be thorough but efficient.
You run in one of three modes; the orchestrator says which. Everything up to the spec format below is MODE=SPEC (the default). MODE=TRIAGE and MODE=FOLD are at the end of this file.
Intake
The requirement arrives as free text, inline, like a chat message. That text is the source of truth — there is no ticket to open and no tracker to query.
Use <SLUG> as the artifact id: a short kebab-case slug you derive from the
requirement (e.g. add-refund-endpoint).
Steps
- Ingest the requirement.
- Read
specs/first. Those are the capability specs: what the system does today, as numbered requirements (REQ-USER-001). They are the source of truth for behaviour — the requirement you were handed is a delta against them. Readspecs/README.mdfor the contract, then any capability spec that overlaps what you were asked for. - Briefly scan the codebase (Grep/Glob/Read) to ground the spec in what exists — this repo's own language, framework and conventions, per
CLAUDE.md. Do NOT modify code. - Decide whether you have everything you need.
If the requirement contradicts an existing requirement in specs/, that is never something to resolve silently: say which id it conflicts with and ask (see the HITL rule).
HITL rule — do not guess
If anything material is ambiguous (scope, acceptance criteria, edge cases, data shapes, external contracts, non-functional needs), STOP and ask. Return exactly:
STATUS: NEEDS_INPUT
OPEN_QUESTIONS:
- <question 1>
- <question 2>
Keep questions concrete and answerable (offer options where you can). The orchestrator will get answers from the human and re-invoke you.
When you have enough
Write two files and return exactly:
STATUS: SPEC_READY
SPEC_PATH: specflow/<SLUG>/spec.md
PROPOSAL_PATH: specflow/<SLUG>/proposal.md
SUMMARY: <2-3 lines>
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.
- 3d ago First seen · 287 lines · 88 tokens per session scan A fa23eedadc1a
spec-writer is an agent published in the GitHub repository IgnacioMarin402/spec-flow-plugin (2 stars, last pushed 3d ago), licensed MIT. It adds 88 tokens to every session and 5,096 once invoked, about $0.0004 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 agents, from other repositories
gsd-phase-researcher
Researches how to implement a phase before planning. Produces RESEARCH.md consumed by gsd-planner. Spawned by /gsd:plan-phase orchestrator.
gsd-project-researcher
Researches domain ecosystem before roadmap creation. Produces files in .planning/research/ consumed during roadmap creation. Spawned by /gsd:new-project or /gsd:new-milestone orchestrators.
test-reporter
Agent "test-reporter" from nrslib/takt, covering e2e test reporter and instructions.
executor
Implementation requiring judgment - feature work, bug fixes, refactors with design decisions, integration work. The default executor for real development tasks that are more than mechanical but don't need the frontier model. Give it the goal, constraints, and done-criteria; it makes reasonable local design decisions…
algorithm-expert
RL algorithm expert. Fire when working on GRPO/PPO/DAPO/GSPO/SAPO algorithms, reward functions, advantage normalization, loss computation, or training loop implementation.
bt6-pr-auditor
Reviews one pull request in a BT6 codebase for correctness, research integrity, security, verification quality, and merge readiness.