AgentSkillsCN

he-intake

通过单一的 slug 和可衡量的成功标准，将模糊的需求转化为具体的行动方案规范，收录于 docs/specs 中。适用于开展非 trivial 任务之初使用。

SKILL.md
--- frontmatter
name: he-intake
description: Converts fuzzy requests into a concrete initiative spec in docs/specs using a single slug and measurable success criteria. Use at the start of non-trivial work.
argument-hint: "[initiative description]"

HE Intake

Create a decision-ready spec artifact for a new initiative.

Output

  • •docs/specs/<slug>.md

Slug

  • •Create one slug: YYYY-MM-DD-kebab-topic
  • •Use this slug for all subsequent phase artifacts

Subagent Usage

When scoping a new initiative, use subagents to research the codebase in parallel — e.g., one subagent to find relevant files and existing patterns, another to check for related specs or prior work in docs/specs/ and docs/plans/completed/. Feed subagent findings into the spec rather than doing all exploration in the main thread.

Intake Procedure

  1. •Define a concise goal statement (problem + target user/outcome).
  2. •Define scope with In Scope and explicit Boundaries (deliberate exclusions).
  3. •Define requirements in a table with stable IDs (R1, R2, ...) and priorities.
  4. •Define measurable success criteria.
  5. •Define constraints (time, risk, compatibility, performance).
  6. •Classify overall priority (critical, high, medium, low).
  7. •Draft initial task graph candidates with rough dependencies using sequence IDs (1, 1.1, 1.2) and include priority for each task.
  8. •Add Next Steps and initialize Change Log.

Progressive Disclosure Rules

  • •Always include: Goal, Scope, Requirements, Success Criteria, Overall Priority, Initial Task Candidates, Next Steps.
  • •Include only when needed: Chosen Direction, Alternatives Considered, Key Decisions, Open Questions.
  • •Keep implementation detail out of intake spec (libraries/endpoints/schema details belong in planning).

Plan Path Selection (Required)

Select plan_mode in the spec metadata:

  • •lightweight when all are true:
    • •Change is small (typically <= 3 tasks and <= 3 files)
    • •No schema migration, auth change, security boundary change, or public API contract break
    • •Overall risk is not critical
    • •Verification can be covered by a short targeted test set
  • •execution for all other work

Write plan_mode: <lightweight|execution> in docs/specs/<slug>.md.

Set spike_recommended: yes if the fuzzy-idea loop concludes a spike is needed, otherwise spike_recommended: no.

Fuzzy-Idea Loop

If request is unclear:

  1. •Capture intended outcome in one sentence.
  2. •Offer 2-3 possible approaches with tradeoffs.
  3. •Pick a default approach and list assumptions.
  4. •If still ambiguous, recommend he-spike before planning.

Spec Template

Use templates/spec-template.md.

Exit Gate

  • •Spec exists at docs/specs/<slug>.md
  • •Goal is explicit
  • •Scope includes explicit boundaries
  • •Requirements table exists with stable requirement IDs (R1+)
  • •Success criteria are measurable
  • •Next steps are explicit
  • •Priority assigned
  • •plan_mode is assigned (lightweight or execution)
  • •spike_recommended is assigned (yes or no)
  • •Docs commit gate passes

Transition Options

Present 2-3 explicit next-step options with a recommended default. Use request_user_input (Codex) or AskUserQuestion (Claude Code) in Plan mode; otherwise ask in chat. Wait for user selection before proceeding.

At least one option must be Next step: he-plan (or Next step: he-spike if a spike is recommended).