Claude-skill-registry agent-ops-impl-details
Extract, plan, or propose implementation details at configurable depth levels (low/normal/extensive). Outputs to reference files for team discussion and handoff.
install
source · Clone the upstream repo
git clone https://github.com/majiayu000/claude-skill-registry
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/majiayu000/claude-skill-registry "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/data/agent-ops-impl-details" ~/.claude/skills/majiayu000-claude-skill-registry-agent-ops-impl-details && rm -rf "$T"
manifest:
skills/data/agent-ops-impl-details/SKILL.mdsource content
Implementation Details
Generate detailed implementation specifications at varying depths.
Reference: See REFERENCE.md for templates, examples, and output formats.
Depth Levels
| Level | Name | Includes | Use Case |
|---|---|---|---|
| 1 | | Files, approach, dependencies, risks | Quick alignment |
| 2 | | + Signatures, pseudo-code, data structures | Technical review |
| 3 | | + Actual executable code, edge cases, tests | Formal spec |
Auto-Selection by Confidence
| Confidence | Level | Override |
|---|---|---|
| LOW | extensive (mandatory) | No |
| NORMAL | normal | Yes |
| HIGH | low | Yes |
LOW confidence requires extensive details with actual executable code.
Modes
Extract Mode
Document existing code to understand how it works.
Input: File path or symbol name
Procedure:
- Identify target (file or symbol)
- Analyze at requested level
- Generate reference file:
.agent/issues/references/{context}-impl-extract.md
Plan Mode
Generate implementation details for planned changes.
Input: Issue ID (e.g.,
FEAT-0083@c5f3e7)
Procedure:
- Read issue requirements and acceptance criteria
- Analyze affected code
- Generate details at requested level
- Create reference file:
.agent/issues/references/{ISSUE-ID}-impl-plan.md - Link from issue's
fieldspec_file
Propose Mode
Suggest implementation approach for a conceptual change.
Input: Description of desired change
Procedure:
- Identify affected areas
- Generate approach options
- Create proposal file:
.agent/issues/references/{context}-impl-proposal.md
Output Sections
All levels include:
- Summary: One-paragraph overview
- Files Affected: Table of files and actions
- Approach: Technical rationale
Normal adds:
- Detailed Changes: Function signatures, code locations
- Dependencies: What this requires/enables
- Risks: Potential issues with mitigations
Extensive adds:
- Complete Implementation: Full executable code
- Edge Cases: Table of cases and handling
- Test Cases: Complete test functions
Invocation
# Extract existing code /agent-impl-details extract src/services/auth.ts --level normal # Plan for issue /agent-impl-details plan FEAT-0083@c5f3e7 --level extensive # Propose approach /agent-impl-details propose "Add rate limiting to API" --level low
Confidence Indicators
Output should include:
## Confidence Indicators ### ✅ High Confidence - All files identified - Clear API contracts ### ⚠️ Uncertainty - {Area}: {Why} ### ❓ Blockers - {Question needing answer}