跳到正文
cccskills
免费在 GitHub 公开

council

Use when the user asks for a council, second opinions, or parallel investigation. Not for cross-examination of existing proposals: use cross-critique.

中文简介正在准备中。当前显示原文说明。

查看安装方法

包含的文件(2)

  • SKILL.md3.6 KB
  • agents/openai.yaml142 B

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

Council

Contract

FieldBound contract
TriggerUser asks for a council, second opinions, or parallel investigation.
AuthorityRead-only. No file, VCS, credential, paid, published, deployed, or remote mutation. Subagents inherit the same read-only posture unless the user explicitly scopes a narrower mutation.
Side effectSpawns parallel subagents in chat output. No repo mutation unless the user explicitly scopes it. Deliverable is the council's decision memo.
DoneDecision memo with recommendation, consensus, risks, and evidence.

Inputs

The question or decision to investigate (required). Number of council members or specific investigative stances to represent (optional; default three distinct stances). Scope constraints and admissible evidence sources (optional). The user-stated criteria bound the investigation; do not widen scope beyond them.

Procedure

  1. Restate the question and confirm scope against the user-stated criteria. Bound the investigation before spawning; do not widen scope. Done when: the question is restated and scope is confirmed against user criteria.
  2. Define council members as distinct investigative stances appropriate to the question (for example optimist, skeptic, risk-auditor, domain expert). Use the user-specified stances when supplied; otherwise default to three. Done when: each council member is assigned a distinct investigative stance.
  3. Spawn one subagent per stance. Each subagent investigates the question independently from its assigned stance, gathers evidence, and returns a position with supporting evidence and identified risks. Done when: one subagent is spawned per stance and each returns a position with evidence and risks.
  4. Collect all subagent positions without merging prematurely. Done when: every subagent position is collected unmerged.
  5. Compare positions across stances: mark consensus points, disagreements, and unresolved risks. Done when: consensus points, disagreements, and unresolved risks are marked across all positions.
  6. Synthesize a decision memo stating the recommendation, the consensus reached, dissenting views, residual risks, and the evidence supporting each. Done when: the memo states recommendation, consensus, dissent, residual risks, and per-point evidence.
  7. Present the memo. Do not select or execute a decision unless the user explicitly authorizes further action. Done when: the memo is presented and no decision is executed without explicit user authorization.

Failure and recovery

  • Subagent returns no usable position: mark that stance as inconclusive, proceed with remaining members, and note the gap in the memo.
  • All subagents inconclusive or evidence insufficient: return a non-converged memo stating the blocker; do not fabricate consensus or pretend the done predicate holds.
  • Scope drift detected: stop, restate the user-stated criteria, and re-bound before continuing; never widen scope to manufacture a result.
  • Subagent attempts repo mutation without user-scoped authorization: treat as out-of-scope, exclude the action, and record the violation in the memo.
  • Partial-result rule: a memo with some inconclusive stances is valid as long as every gap is explicitly recorded; never silently drop a stance.

Output

A decision memo containing recommendation, consensus, dissenting views, residual risks, and evidence. Terminal classification is converged (consensus reached) or non-converged (insufficient evidence or unresolved disagreement).

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

Use when explaining System V AMD64, ARM AAPCS, RISC-V psABI, stack frames, variadic calls, or FFI register rules. Not for the Rust FFI binding layer: use rust-ffi.

中文简介正在准备中。当前显示原文说明。

OutlineDriven/odin-claude-plugin372026年9月29日 更新

Use when configuring ADC sampling time, DMA-driven ADC, calibration, or DAC channel setup on bare-metal MCUs. Not for the DMA stream itself: use dma-baremetal.

中文简介正在准备中。当前显示原文说明。

OutlineDriven/odin-claude-plugin372026年9月29日 更新

af-xdp

免费

Use when creating AF_XDP sockets, configuring UMEM and XSK rings, writing an XDP redirect program, or choosing copy versus zero-copy mode. Not for full kernel bypass: use dpdk.

中文简介正在准备中。当前显示原文说明。

OutlineDriven/odin-claude-plugin372026年9月29日 更新

Use when a completed session needs an agent-environment retrospective. Not for an engineering retrospective from telemetry: use engineering-retrospective.

中文简介正在准备中。当前显示原文说明。

OutlineDriven/odin-claude-plugin372026年9月29日 更新

Use when a redacted, trimmed agent transcript must be appended to a GitHub PR or issue body, with human approval and preview. Not for automated or model-initiated insertion.

中文简介正在准备中。当前显示原文说明。

OutlineDriven/odin-claude-plugin372026年9月29日 更新

agents-md

免费

Use when a repo needs agent setup, AGENTS.md added or made lean, CLAUDE.md audited, or agent instructions scored or pruned. Not for remote, credential, publish, deploy, or irreversible changes.

中文简介正在准备中。当前显示原文说明。

OutlineDriven/odin-claude-plugin372026年9月29日 更新

查看 OutlineDriven 的全部技能

报告此技能的问题