code-teardown

A report that studies an external code repository—such as an open-source project or competitor’s code—to identify ideas your project could adopt. Each finding includes supporting evidence from the repository.

In plain words
What is it for?
Use it to compare approaches, evaluate a trending project, and find adoptable architecture, coding patterns, developer-tooling, testing, and documentation ideas.
Why use it?
It turns a large unfamiliar codebase into specific, evidence-backed decisions instead of leaving you with a general architecture tour.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/lucassantana-dev/sharekit/code-teardown
Any agent
npx skills add LucasSantana-Dev/sharekit --skill code-teardown
Clone the repo
git clone --depth 1 https://github.com/LucasSantana-Dev/sharekit

Made for: Claude Code, Codex.

Per session 168 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,259 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00168 $0.01259
Opus 5 $0.00084 $0.00629
Sonnet 5 $0.00034 $0.00252
Haiku 4.5 $0.00017 $0.00126

Measured 2d ago against content hash 65fa35b2336e, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

code-teardown scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

sharekit-profile/.claude/skills/code-teardown/SKILL.md · 98 lines

How it starts

The opening of the file, as written. The whole thing — 98 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Code Teardown

Reads someone else's codebase to answer one question: what do they do that WE should do? Same contract as app-teardown (live apps); this one targets repos and reads code. Output is a decision artifact with verdicts, not an architecture tour — for "how does X work" documentation use doc-and-modernize.

Use When

  • An external repo should be mined for architecture/pattern/DX/tooling ideas for one of our projects.
  • User names a repo, OSS project, or codebase and asks what to learn, adopt, or steal from it.
  • Evaluating a trending project ("should we care about X?") — the negative outcome is captured too.

Do Not Use When

  • Target is a live product to browse → app-teardown.
  • You will contribute to or modify the target repo → adt-repo-intake.
  • Documenting OUR OWN system → doc-and-modernize.

Inputs / Prereqs

  • Target repo (URL or local path). Clone shallow to ${DEV_ROOT}/<repo> (storage policy; never $HOME, never internal disk).
  • OUR context lock: which project the ideas land in, its stack and constraints.
  • Memory state-check FIRST: reference_<target>_evaluated_* exists → surface prior outcome and stop unless the revisit condition is met.
  • Target is read-only: no issues, no PRs, no edits there.

Workflow

  1. Scope + constraint capture — why does their design work FOR THEM (scale, team size, ecosystem, age, funding model)? Record upfront; feeds the cargo-cult gate.
  2. Shallow clone + intake sweepgit clone --depth 1. Read README, docs/, ADRs, CONTRIBUTING, CI config, package manifest, top-level layout. For large repos dispatch read-only Explore agents per area instead of reading serially.
  3. Map by dimension — Architecture (boundaries, data flow, state), Code patterns (idioms, error handling, abstractions worth naming), DX (scripts, codegen, local-dev loop), Tooling/CI (build, release, quality gates), Testing (strategy, fixtures, ratio), Docs conventions (ADRs, changelogs, onboarding). 2+ dimensions minimum; note deliberate ABSENCES (what they chose not to build).
  4. State-check per candidate finding — do WE already have it? graphify query on the target project (if graphify-out/ exists) + grep + memory. Hits → already-have.
  5. Verdict + rationale per finding — apply the shared contract: ~/.agents/skills/app-teardown/references/teardown-contract.md (evidence as file:line at a pinned commit SHA, verdict, mandatory "[their constraint] → [applies to us?] → [action]" rationale, effort, landing spot).
  6. Emit + remember — report to <project>/docs/reference/<target>-teardown-<date>.md. Memory note reference_<target>_evaluated_<date>.md with outcome + revisit condition; "evaluated → nothing" is a first-class outcome (megabrain/llmwiki precedent). Summary: "N findings, A adopt, D adapt, H already-have, R rejected". Remove the clone if nothing was adopted and no revisit is planned.

Read the full file on GitHub · 98 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 98 lines · 168 tokens per session scan A 65fa35b2336e

Subscribe to this mod's changes

code-teardown is a skill published in the GitHub repository LucasSantana-Dev/sharekit (1 stars, last pushed 2d ago), licensed MIT. It adds 168 tokens to every session and 1,259 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.