AGENT SKILL

techdebt

Inventory tech debt across the codebase - duplicated logic, dead code, magic numbers, deep nesting, TODO/FIXME rot - as a report of locations + suggested actions, not auto-applied fixes. Use when you want a broad debt sweep. NOT for applying fixes to just-changed code (use the built-in /simplify) and NOT for reviewing a specific diff or PR (use /code-review).

License 许可
MIT
所属仓库
SamyakJhaveri/loam
最近更新
2026-11-15
Stars
★ 0
Forks
0

01 前言 Front Matter

name
techdebt
model
claude-opus-4-8[1m]
description
Inventory tech debt across the codebase - duplicated logic, dead code, magic numbers, deep nesting, TODO/FIXME rot - as a report of locations + suggested actions, not auto-applied fixes. Use when you want a broad debt sweep. NOT for applying fixes to just-changed code (use the built-in /simplify) and NOT for reviewing a specific diff or PR (use /code-review).
allowed-tools
[ "Read", "Grep", "Glob", "Bash(*)" ]

03 SKILL.md 正文

Tech-debt inventory

Produce a report of tech debt - locations + suggested actions, not auto-applied fixes. (For fixes to just-changed code use the built-in /simplify.)

Scope

Default to a whole-tree sweep - this is the broad-inventory skill. To scope to recent work instead, use git diff --name-only main...HEAD. Do NOT run git diff --name-only HEAD~5 blindly: it errors on a branch with fewer than 5 commits - fall back to main...HEAD, then to a whole-tree scan.

Scan - one pass per debt class

  1. Duplicated logic - Grep for repeated blocks/signatures; propose a shared helper.
  2. Dead code - Glob the module tree; Grep each unused-looking import/function to confirm zero references before flagging.
  3. Magic numbers/strings - Grep for bare literals in logic; propose named constants.
  4. Deep nesting (>3 levels) - flag for an early-return refactor.
  5. TODO/FIXME/XXX rot - list (with age if cheap); never auto-fix.

Output - one table

Location (file:line) Class Suggested action

This skill reports; it does not edit. Hand the table back to the user and let them choose what to act on. If they approve a set of fixes, apply them outside this skill, then run the project's linter/type checker (/validate) to confirm no regressions.

04 同仓库更多 Skills

SamyakJhaveri
SamyakJhaveri/loam

verify-loam

Prove a Loam change in a real rendered project: copier-render this checkout, then drive headless and interactive Claude Code sessions, stubbed Codex runs and copier updates in the render and capture hook, memory, deny-rule, update and check evidence. Use before shipping any seed, hook, settings or copier.yml change.

★ 0
SamyakJhaveri
SamyakJhaveri/loam

hypothesis-tree

Persistent hypothesis tree for multi-step, multi-session investigations. Use when a "why does X fail" question branches into sub-hypotheses that must survive across sessions, or when falsifiable claims need tracking with dated, linked evidence. Maintains one markdown file with add, update, review, and prune subcommands, each gated on verifying the cited evidence file exists and supports the claim. NOT for a single-session bug hunt, a one-command question, or general note-taking.

★ 0
SamyakJhaveri
SamyakJhaveri/loam

fable-prompting

Which Claude Fable 5.2 guide sections you can act on from a session, and which Claude Code already injects. Use before writing a Fable session prompt or brief.

★ 0
SamyakJhaveri
SamyakJhaveri/loam

catchup

Read-only session bootstrap briefing of git state, commits, handoff, tasks and red flags. Use when resuming work, starting a fresh session, or unsure of project state.

★ 0
SamyakJhaveri
SamyakJhaveri/loam

writing-for-agents

Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.

★ 0
SamyakJhaveri
SamyakJhaveri/loam

wizard

Generate an interactive bash wizard that walks a human through steps only they can perform. Use when provisioning infrastructure, setting up credentials or CI secrets, walking an unfamiliar third-party dashboard, or running a one-off migration or cutover. Don't invoke this for steps the agent can perform itself.

★ 0