菜单
Diagnose a concrete failure with rival-hypothesis testing. Stabilize the symptom, generate distinct explanations, test discriminating observations read-only in parallel, and rank by evidence while retaining losing rivals.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/m0n0x41d/haft --skill h-diagnose命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
Source-first umbrella for FPF-aware reasoning in a Haft project. Use for ambiguous engineering, management, architecture, specification, or project questions when no narrower Haft capability is already current. Ordinary reasoning stays conversational; persistence is conditional and binding actions remain manual.
Source-first umbrella for FPF-aware reasoning in a Haft project. Use when the operator asks to think through an ambiguous engineering, management, architecture, specification, or project question without naming a narrower Haft capability. Recover the current object and question, query the bundled FPF source, inspect the governing pattern body, and choose only the capability that is current. Ordinary reasoning stays conversational; proactively persist the minimum needed memory when current Work supplies a concrete durable receiving use. h-decide may route a direct operator request; h-commission remains manual-only.
Creates a WorkCommission — bounded execution authority — from an active DecisionRecord. MANUAL ONLY: operator must explicitly type /h-commission. Never auto-invoked: commissions are execution-authority grants under Transformer Mandate. Runs freshness check, scope check, derives an ImplementationPlan, snapshots the autonomy envelope, then STOPS before execution. NOT for the decision itself (use /h-decide first). NOT for running tests or one-off tasks (the operator's coding agent handles those directly).
基于 SOC 职业分类
| name | h-diagnose |
| description | Diagnose a concrete failure with rival-hypothesis testing. Stabilize the symptom, generate distinct explanations, test discriminating observations read-only in parallel, and rank by evidence while retaining losing rivals. |
Retrieve current FPF source only when pattern choice is material to the
diagnosis. Mechanical reproduction, logs, and direct implementation probes do
not need a ritual query. Inspect a known SourceID/UnitID directly; otherwise
use mode="concern" and treat the returned candidate_set as incomplete
navigation, not a selected pattern. Before relying on one candidate, inspect
its exact identifier and direct pattern body. Keep several candidates live or
abstain when the returned basis is insufficient. Never run a query after the
diagnosis merely to manufacture source support for an already-chosen story.
Keep symptom, hypothesis, probe, observation, and verdict distinct. Include a rival that challenges the initial framing. Prefer safe parallel probes; label design-time inference separately from runtime evidence. Keep losing hypotheses with return conditions. Persist only on explicit request or when a named receiving use needs replay.
