What is repo-conventions?

A repository-specific review guide containing NeuroLink's engineering rules and the issues reviewers should focus on.

Use it when reviewing changes in the NeuroLink repository, especially provider imports, tool schemas, and other documented critical rules.

It keeps code reviews aligned with the project's own standards and avoids repeating problems already checked automatically by CI.

repo-conventions — Skill for Claude Code and Codex by juspay/neurolink

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.

Any agent

npx skills add juspay/neurolink --skill repo-conventions

Clone the repo

git clone --depth 1 https://github.com/juspay/neurolink

Made for: Claude Code, Codex.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/juspay/neurolink/repo-conventions/github.svg)](https://agentmods.dev/skills/juspay/neurolink/repo-conventions)

Your own site

<a href="https://agentmods.dev/skills/juspay/neurolink/repo-conventions"><img src="https://agentmods.dev/badge/skills/juspay/neurolink/repo-conventions/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.

Your own site · 80×15

<a href="https://agentmods.dev/skills/juspay/neurolink/repo-conventions"><img src="https://agentmods.dev/badge/skills/juspay/neurolink/repo-conventions.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 39 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,210 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits

NVIDIA SkillSpector pass 7 Sept 2026

How audits are shown

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 5.1	$0.00039	$0.01210
Opus 5.5	$0.00016	$0.00484
Sonnet 5.5	$0.00008	$0.00242
Haiku 4.5	$0.00004	$0.00121

Measured 1mo ago against content hash a29da7cd54bd, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

repo-conventions 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 1mo 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.

.yama/skills/repo-conventions/SKILL.md · 96 lines

How it starts

The opening of the file, as written. The whole thing — 96 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Reviewing a change in NeuroLink

Authoritative sources in this checkout (the repository root is ../):

../CLAUDE.md — the engineering rules, long form. When a finding violates one, cite the specific rule (e.g. "Critical Rule 1: dynamic imports only in the registry").
../CONTRIBUTING.md — contribution and commit conventions.

This skill is the review-focused digest; those files win on any conflict.

Do NOT comment on — CI already owns it

Formatting, lint and type errors — including CLAUDE.md rules 2 and 7–15 — are enforced by ESLint + Prettier + tsc and the custom AST rules in eslint-rules/ (interface vs type, type file locations, barrel rules, double assertions, e2e-only tests). Re-reporting mechanical violations buries the findings that matter. Also skip: performance micro-costs, naming and prose taste, dependency choice.

The critical rules the review DOES enforce (cite the rule number)

Rule 1 — dynamic imports only in the registry. Providers are imported dynamically inside factory functions in providerRegistry.ts; a static provider import is a circular-dependency bug.
Rule 3 — Gemini tools + JSON-schema structured output are mutually exclusive. Gated on isGeminiProvider in structuredOutputPolicy.ts, NOT on Vertex as a whole (Vertex Claude supports both). Claude paths must default max_tokens via resolveClaudeMaxTokens, never a hardcoded 4096; truncation must surface (jsonRepaired / jsonTruncated), never silently.
Rule 4 — CLI ≠ SDK. Manual MCP connections are CLI-only; CLI concerns must not leak into the SDK path.
Rule 5 — backward compatibility of the public SDK API. Non-negotiable. Name the unmodified callers a change breaks; use the code graph to find them.
Rule 6 — formatProviderError returns the error, never throws.
Rule 15 — tests are end-to-end only. Suites drive dist/index.js or the built CLI (node dist/cli/index.js); ONE module graph per suite — mixing src/ and dist/ imports breaks stubs/spies/instanceof silently. A determinism exception must be declared in the file header and in the ESLint allow list.
Provider executeStream goes through BaseProvider.stream() tool-merge.
Factory + Registry is the extension pattern (providers, processors, chunkers, rerankers) — a new extensible thing that bypasses it is a finding.

Read the full file on GitHub · 96 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.

1mo ago First seen · 96 lines · 39 tokens per session scan A a29da7cd54bd

Subscribe to this mod's changes

repo-conventions is a skill published in the GitHub repository juspay/neurolink (157 stars, last pushed today), licensed MIT. It adds 39 tokens to every session and 1,210 once invoked, about $0.0002 per session on Opus 5.5. 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.

Code review LLM agents ai developer-tools mcp model-context-protocol

gh-address-comments

This skill should be used when addressing GitHub pull request review comments systematically with mandatory resolution tracking. Use when user says "address PR comments", "fix review feedback", "handle PR review", "resolve PR threads", or needs to ensure all review comments are addressed before merging. Fetches…

jmagar/claude-homelab · 85 tokens

repo-conventions

Getting it into your agent

Wrote this? Show the measurements

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

Reviewing a change in NeuroLink

Do NOT comment on — CI already owns it

The critical rules the review DOES enforce (cite the rule number)

What this file has done since we first saw it

Other skills, from other repositories

gh-address-comments

otto-pr-review-loop

resolve-drive-pr

code-review

resolve-review-pr

describe-pr