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 0fuk/OpenACCP --skill need-researchgit clone --depth 1 https://github.com/0fuk/OpenACCPWrote 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/0fuk/openaccp/need-research)<a href="https://agentmods.dev/skills/0fuk/openaccp/need-research"><img src="https://agentmods.dev/badge/skills/0fuk/openaccp/need-research/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/0fuk/openaccp/need-research"><img src="https://agentmods.dev/badge/skills/0fuk/openaccp/need-research.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 | $0.00124 | $0.01254 |
| Opus 5.5 | $0.00050 | $0.00502 |
| Sonnet 5.5 | $0.00025 | $0.00251 |
| Haiku 4.5 | $0.00012 | $0.00125 |
Grade A, and why
need-research 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 2d 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.
What it actually says
/need-research — 决策级盲区调研
语义
用户潜台词:「我要做决策,但怀疑认知有洞。」交付三层:①你问的(证据齐全)②你没问但该知道的(按咬人程度排序)③决策选择题(每题带推荐默认值+一句理由;拍板权在用户)。
硬约束
- 总代理 ≤10;常态 3-8;宽度跟真实维度走,禁凑数。
- 全程禁 opus;主循环(定题/盲区地图/仲裁/终审/写报告)零子代理;所有
agent()显式 model+effort。 - 调研=sonnet·high;挑刺复核=sonnet·max 双盲;机械盘点=haiku·low(大切片改 sonnet)。
- 带预算守卫;截断必 log 被放弃的覆盖面。
流程
第 0 步 定题(主循环亲做):①先定这次调研服务哪个决策——没决策的调研是消遣;不清楚就问(≤3 问);无人值守窗改为按默认假设开工+入候 owner 清单。②已知盘点——先查 ~/.claude/research-index.md 与项目文档有无同题既往报告,只研增量。③画盲区地图——从原题出发列相邻维度:风险合规/真实全周期成本/替代方案(含不做)/时间窗/退出成本/二阶后果/内行会追问什么。④维度清单 3-8 条,至少 1 条纯盲区猎手。
第 1 步 调研并发派出:每维度一代理。prompt 必含证据纪律(结论带来源+日期;营销页标 [营销源] 降权;找不到写找不到);一手优先;findings 落盘 scratchpad,返回结构化 key_claims(id/claim/evidence/confidence)。盲区猎手 prompt 明写「任务不是答 X,是找 X 周边用户没意识到的坑」。
第 2 步 双盲对抗(1-2 个,互不知晓):V1 证据实在性(逐条重抓来源,不支撑=REFUTED/弱=WEAKENED);V2 反证搜索(换关键词/语言/信源找反例,专抓过期信息、促销价冒充常价、幸存者偏差)。另答 missed:整个调研漏了什么。
第 3 步 终审(主循环):仲裁 REFUTED;冲突亲开证据裁决;missed 重要的补 1 代理或诚实标未覆盖;写报告;证伪战果单独成节。两轮挑刺复核零 REFUTED/WEAKENED 时主循环亲抽 2-3 条承重 claim。
交付规格
- 最终报告 HTML(不是 md),浅色、说人话;视觉引用项目设计规范(无则 orch-plan 的通用骨架)。
- 结构(顺序即优先级):一句话答案 / 你问的 / 你没问但该知道的 / 证伪战果 / 决策选择题 / 诚实未知 / 证据附录(可折叠)。
- 决策选择题格式与 /feature-dev 决策包统一:每题=选项+推荐默认值(加粗)+一句理由+代价与可逆性(选错后果/何时还能反悔)。不替用户拍板。
- 落盘报完整绝对路径;同时在
~/.claude/research-index.md追加一行(题目+路径+日期),供下次「只研增量」。 - 排期建议禁日历日串行,画依赖图。
规模路由(报给用户选了哪档)
轻(单维度核查)=2 研+1 挑刺复核;标准(选型/方案比较)=4-5 研(含 1 猎手)+2 挑刺复核;重(战略/合规/大额)=6-8 研+2 挑刺复核(硬顶 10)。超出重档分两轮,第一轮收敛后由用户定第二轮方向。
反模式
只答原题不扫盲区;盲区清单凑数(每条必答「不知道会怎么被咬」);opus「就这一个环节」;建议写成论文而非选择题;挑刺复核只盖章不证伪;有既往报告还整题重跑;报告落成 md。
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.
- 2d ago First seen · 35 lines · 124 tokens per session scan A 234d428fa8b0
need-research is a skill published in the GitHub repository 0fuk/OpenACCP (2 stars, last pushed 2d ago), licensed MIT. It adds 124 tokens to every session and 1,254 once invoked, about $0.0005 per session on Opus 5.5. 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 2026-10-14.
Other skills, from other repositories
handoff-mesh
Built-in communication layer for multi-agent setups. Full-mesh handoff folders (any agent can hand off to any other), an hourly watcher that surfaces open handoffs, weekly archive cleanup, and a completion-reporting loop so every handoff is handled AND verified. The durable 'brain' counterpart to real-time channels…
agent-teams
Coordinates Claude agent teams via filesystem protocol. Use when orchestrating parallel agents with task dependencies. Do not use for single-agent tasks.
orchestrate
This skill should be used when the user asks to 'orchestrate a task', 'break down work into parallel agents', 'coordinate subtasks', 'run agents in parallel', or mentions 'multi-agent'. Decomposes complex tasks into tracked subtasks, dispatches parallel subagents, and coordinates until completion.
silicon-steward
A planning guide for coordinating work between a person and a coding agent. It turns required information, resources, feedback, actions, and handoffs into explicit checklists and protocols.
coordinate-recover
Diagnose and safely resume Coordinate Agents Tasks after executable failure, non-zero exit, timeout, stale claim, processing message, or Implementer ERROR. Recovery is explicit and never an automatic retry loop.
coordinate-setup
Discover or configure implementation agents for Coordinate Agents when setting up or changing its agent configuration.