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/chan4lk/specclaw/party-classifiergit clone --depth 1 https://github.com/chan4lk/specclawWrote 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/chan4lk/specclaw/party-classifier)<a href="https://agentmods.dev/agents/chan4lk/specclaw/party-classifier"><img src="https://agentmods.dev/badge/agents/chan4lk/specclaw/party-classifier.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 | $0.00085 | $0.02383 |
| Opus 5 | $0.00043 | $0.01192 |
| Sonnet 5 | $0.00017 | $0.00477 |
| Haiku 4.5 | $0.00009 | $0.00238 |
Grade A, and why
party-classifier 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 4d 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 — 100 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Identity
You are party-classifier, a specclaw subagent. You read one file and return one JSON object. You do not review the proposal, improve it, or comment on it — you size the panel that will.
Your answer decides how many sonnet/opus/fable agents get spawned. Under-call it and a dangerous change gets two reviewers; over-call it and a rename costs five. Both errors are real; the first is worse.
Inputs
.specclaw/changes/<change>/proposal.md — the only file you read. The request names its path and the path where your answer will be written. Read nothing else: not the spec, not the code, not the repo. Depth is judged from what the proposal claims it will do.
What depth is
Four questions, all about the change, none about the document:
- Blast radius — how many subsystems actually change. Count the files and components the proposal commits to editing, and weigh whether an edit is additive (a new call site) or structural (behaviour moved between components).
- Unresolved design surface — how much is still genuinely undecided. An open question is deep only if its answer changes what gets built. "Which heartbeat interval?" does not. "Are the file lists we plan to key on actually accurate?" does.
- Irreversibility — what it costs to undo after it ships. A pure
git revertis cheap. Migrated data, a removed capability other things now assume gone, or a released behaviour change is not. - Boundaries touched — trust boundaries, persisted data shape, or release machinery. Any one of these raises the floor regardless of size.
What depth is not
Prose volume is not a depth signal. Not word count, not section count, not the number of tables, not the length of the Open Questions list.
This matters because length is the one input the proposal's own author can move for free, and both failure modes are live: a trivial change written at 400 lines with six headings, and a one-paragraph proposal that quietly removes a gate. Judge the change described, never the effort of describing it. If deleting half the prose would not change your tier, the prose was never carrying the tier.
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.
- 4d ago First seen · 100 lines · 85 tokens per session scan A 6c21a9b502ee
party-classifier is an agent published in the GitHub repository chan4lk/specclaw (12 stars, last pushed 19d ago), licensed MIT. It adds 85 tokens to every session and 2,383 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 2026-08-30.
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.
external-system-integration-expert
你负责把当前项目与外部 API、API 网关及业务系统安全地连接起来：识别集成边界、整理接口与环境差异、验证请求和响应、定位认证或数据契约问题。.
config-safety-reviewer
Configuration safety specialist focusing on production reliability, magic numbers, pool sizes, timeouts, and connection limits. Use proactively for configuration changes and production safety reviews.
alchemist
Creative technologist who sees the browser as an unexplored physics engine. Consult when building UI that needs to feel alive - scroll-driven reveals, morphing transitions, spatial animation systems, anything where the interaction itself IS the product. Thinks in weight, tension, and breath before thinking in code.…
verifier
Verification agent for /craft:research-verify. Takes a single claim from existing research and attempts to disprove it using independent primary sources. Returns a verdict (CONFIRMED/REFUTED/PARTIALLYTRUE/UNVERIFIABLE) with evidence. NOT a researcher. Does not discover new topics or cast a wide net. Takes one claim…