A coding skill that checks project documentation and locked product decisions before implementing a feature. It also consults design references and follows project coding rules.

Use it to implement approved features, check required documentation, route unresolved design questions, and keep code aligned with the project's specifications.

It prevents implementation from starting when essential project context is missing or important feature choices are still undecided.

code-agent — Skill by sauravvarma

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/sauravvarma/garage/code-agent

Any agent

npx skills add sauravvarma/garage --skill code-agent

Clone the repo

git clone --depth 1 https://github.com/sauravvarma/garage

Made for: Claude Code, Codex.

Per session 113 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 5,987 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.

Model	Per session	Once invoked
Fable 5	$0.00113	$0.05987
Opus 5	$0.00056	$0.02993
Sonnet 5	$0.00023	$0.01197
Haiku 4.5	$0.00011	$0.00599

Measured 2d ago against content hash 68cbfd176790, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

code-agent 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.

skills/code-agent/SKILL.md · 222 lines

How it starts

The opening of the file, as written. The whole thing — 222 lines — stays where its author put it; the contents beside it link to each section on GitHub.

code-agent — coding specialist

Spec preflight

Before doing the work, verify the artifacts this skill depends on. Stop and present a checklist if anything required is missing — do not silently proceed with assumptions.

Required:

CLAUDE.md — project context, commands, preflight section
docs/REPO-CONVENTIONS.md — architectural patterns (data flow, routing, state, code splitting, analytics, styling)
docs/COMPONENT-SPECS.md — code component index and shared atoms

Recommended (for the feature in scope):

docs/[FEATURE]-IDEAS.md — decision table with locked decisions, route chrome, design rationale

Optional:

.claude/rules/ — project-specific gates

If anything required is missing, tell the user exactly which docs are missing and offer two paths:

(default) "Run /spec-first-project-setup to scaffold the missing pieces, then come back."
"Proceed without — I'll make assumptions and flag them as I go."

If [FEATURE]-IDEAS.md exists but has any decision in the table with status open for the feature in scope, refuse to implement until those decisions are locked. Route the user using the lock-state hierarchy:

If docs/DESIGN-HEURISTICS.md is missing (or any open decision row has Type: direction) → direction itself isn't locked. Say: "Decisions [N], [M] are still open and direction-level. Run /design-explore '[feature intent]' to discover and lock the direction first."
If docs/DESIGN-HEURISTICS.md is present and open decisions are all Type: variant (or untyped, in which case assume variant once direction is locked) → variant choices within a locked direction. Say: "Decisions [N], [M] are still open within the locked direction. Run /design-agent '[feature intent]' to generate variants and let the user pick."

(Same lock-state hierarchy as /design-agent's spec preflight — keep both in sync if changing.)

To match the user's task to its [FEATURE]-IDEAS.md, use the closest filename match. If ambiguous or none, ask the user which feature doc applies.

Read the full file on GitHub · 222 lines

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

references/snapshot-hygiene.md 5.1 KB

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.

2d ago First seen · 222 lines · 113 tokens per session scan A 68cbfd176790

Subscribe to this mod's changes

code-agent is a skill published in the GitHub repository sauravvarma/garage (3 stars, last pushed 1mo ago), licensed MIT. It adds 113 tokens to every session and 5,987 once invoked, about $0.0006 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.

Building agents claude-code claude-code-plugin design-system engineering-workflow lifecycle pencil

background

Use when the user wants to see, inspect, cancel, or prune background agents fired during prior chain runs. Read/manage .hyperflow/background/registry.json and the per-agent output buffers at .hyperflow/background/ .md. Standalone — never auto-invoked. Trigger with /hyperflow:background, "list background agents"…

jeremylongshore/tons-of-skills-marketplace · 89 tokens

lingtai-tui-help

Anything you need to know about LingTai TUI. Read this first for the stable mental model, major-feature map, interface and slash-command help, or the correct source/manual route for a TUI question. It stays thin: exact command prose lives in the localized help assets, precise Go structure lives in lingtai-tui-anatomy…

Lingtai-AI/lingtai · 106 tokens

omd:autopilot

One-prompt autonomous product design and implementation. Use automatically for broad greenfield UI requests such as 'from scratch', '새 제품/화면을 알아서 만들어줘', or requests that delegate DESIGN.md creation. It decides whether to reuse, establish, refresh, or skip a project design system; asks at most one consequential…

kwakseongjae/oh-my-design · 99 tokens

code-agent