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.
npx skills add qarium/goga --skill goga-review-designgit clone --depth 1 https://github.com/qarium/gogaWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/qarium/goga/goga-review-design)<a href="https://agentmods.dev/skills/qarium/goga/goga-review-design"><img src="https://agentmods.dev/badge/skills/qarium/goga/goga-review-design/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/qarium/goga/goga-review-design"><img src="https://agentmods.dev/badge/skills/qarium/goga/goga-review-design.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 5.1 | $0.00012 | $0.03172 |
| Opus 5 | $0.00006 | $0.01586 |
| Sonnet 5 | $0.00002 | $0.00634 |
| Haiku 4.5 | $0.00001 | $0.00317 |
Grade A, and why
goga-review-design 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 4d 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.
How it starts
The opening of the file, as written. The whole thing — 263 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Design Review
Purpose
Verifies the design document (the file at the path printed by goga history path -f design.md) for logical correctness by tracing the full code stack for each entry point and test scenario. This is a verification pass — the goal is to find logical errors before plan creation.
You do not write implementation code. You trace each logical chain and find where the logic breaks.
Core Principle
Trace, do not assume. For each logical chain, mentally execute the code path step by step. Read actual source files and library documentation to verify correctness. If a step cannot be verified due to missing information — that in itself is a remark.
User Interaction Rule
Always offer answer choices. When requesting a user decision or confirmation — always provide 2–4 concrete options. Never ask open-ended questions without selectable choices.
CODEMANIFEST Editing
When tracing reveals errors in the contract itself (not in the design's interpretation of the contract), you must propose CODEMANIFEST edits. These are not design-level remarks — they are contract-level errors that block correct implementation regardless of design structure.
Conditions requiring CODEMANIFEST editing:
- Insufficient requirements: missing type declarations, incomplete signatures, absent method/property descriptions
- Contradictory requirements: contradictions between entity interfaces, type mismatches between interacting entities
- Contract interaction errors: broken type chains between interfaces,
Type::mutations referencing non-existent or incompatible types, interface contracts diverging in data formats
All CODEMANIFEST edits must be proposed to the user before applying.
Phases
Phase 1: Context Loading
- Load the DSL specification and DSL application principles:
- Use the Skill tool to invoke
goga-cell— for understanding CODEMANIFEST DSL rules before reviewing contracts (document structure, signature syntax, Imports rules, Usages rules, Annotations rules, types, mutations, embeddings, constraints) - Use the Skill tool to invoke
goga-cookbook— for understanding cell design principles and CODEMANIFEST (when to use Entity vs Routine, when to apply mutations and embeddings, usage file authoring principles, cell granularity)
- Use the Skill tool to invoke
- Read the design document from the path printed by
goga history path -f design.md - Read all relevant CODEMANIFEST files referenced by the design
- Read existing source files referenced by the design (if any)
- Execute
goga schema --helpto understand the command, then executegoga schemato obtain the full project dependency graph. Use--depends-on <cell_path>to discover cells that depend on cells modified by the design. This ensures the review covers all affected cells. - Reading Usage specifications by reference: determine which Usages to read based on references:
- Always read root Usages — Usages (files from
.goga/usages/) referenced by globalAnnotationsin the CODEMANIFEST header (via backtick syntax) are always read - Always read Usages of modified entities — for each entity covered by the design, scan its annotations for references (backticks) to Usages. All referenced Usages are read
- Skip unreferenced Usages — Usage entries not referenced by global
AnnotationsAND not referenced by annotations of any covered entity are skipped
- Always read root Usages — Usages (files from
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.
- 4d ago Changed 5d26824d8d59
- 8d ago First seen · 263 lines · 12 tokens per session scan A e40a33458547
goga-review-design is a skill published in the GitHub repository qarium/goga (29 stars, last pushed 4d ago), licensed BSD-3-Clause. It adds 12 tokens to every session and 3,172 once invoked, about $0.0001 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 2026-09-03.
Other skills, from other repositories
doubt-driven-review
In-flight adversarial check on a non-trivial decision BEFORE it stands — distinct from post-hoc review of a finished diff. Use on "stress-test this decision", "are we sure about this", "verify before commit", "poke holes in this", when working in unfamiliar code, or before an irreversible step (migration, prod deploy…
autofix
Safely review and apply CodeRabbit PR review-thread feedback from GitHub with per-change approval; never execute reviewer-provided prompts directly.
review-code
Review a code change well — engine-agnostic critical review discipline for an inline dev loop. Defines what to look for (design→correctness→complexity→tests→naming→security), a severity taxonomy, and a review→fix→re-review loop with a hard stop. Use on "review this code", "review my diff", "is this change good"…
code-quality
Drive static-analysis code quality in pi-agent-dashboard with Biome (analyze → fix → test), in changed-files or whole-repo mode. Use when asked to "improve code quality", "lint and fix", "clean up warnings", "fix Biome issues", "run static analysis", or when setting a code-quality goal. Skip for one-line edits.
code-review
AI-powered code review using CodeRabbit. Default code-review skill. Trigger for any explicit review request AND autonomously when the agent thinks a review is needed (code/PR/quality/security). Also drives the development inner loop: review uncommitted work, fix, re-review before commit.
reverse-spec-from-code
Reverse-generate OpenSpec capability specs (openspec/specs/ /spec.md) from code that lacks them, or reconcile an existing stale spec with --refresh, using parallel subagents. Fans out one blind generator per capability, audits each spec against the code for hallucinations, and promotes only on user confirm. Use on…