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 agentmods add instructions/wuzf/claude-channels-patch/claude-mdgit clone --depth 1 https://github.com/wuzf/claude-channels-patchWrote 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/instructions/wuzf/claude-channels-patch/claude-md)<a href="https://agentmods.dev/instructions/wuzf/claude-channels-patch/claude-md"><img src="https://agentmods.dev/badge/instructions/wuzf/claude-channels-patch/claude-md.svg" alt="Measured on agentmods" 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 | $0.00482 | $0.00482 |
| Opus 5 | $0.00241 | $0.00241 |
| Sonnet 5 | $0.00096 | $0.00096 |
| Haiku 4.5 | $0.00048 | $0.00048 |
Grade A, and why
claude-channels-patch CLAUDE.md 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 5d 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.
What it actually says
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
Project Overview
Binary patch for Claude Code that enables the --channels feature without requiring claude.ai OAuth authentication. Patches three gates: feature flag (tengu_harbor), OAuth check (accessToken), and channel allowlist.
Architecture
Python patch script (patch.py) that:
- Auto-detects all installed Claude Code binaries (native
.exe+ npm versions) - Finds stable anchors — string literals and property names that don't change between builds (e.g.,
.marketplace))return{action:"skip",kind:"allowlist",tengu_harbor",!1)}) - Searches backwards from anchors to locate the exact byte to change (always
!-> space or1->0) - Single-byte replacements — every edit changes exactly 1 byte, keeping binary size unchanged
The Claude Code binary is a Node.js SEA with two copies of the JS bundle, so every patch must be applied twice.
Key Design Constraints
- No minified name dependency: Variable names like
SL,D,OaHchange between builds. All anchors use only stable strings (return values, property names, string literals). - Equal-length replacement: All patches are 1-byte changes. No shifting, no size change.
- Backup-first: Always patches from the
.bakcopy, never from an already-patched binary. - Cross-platform: Pure Python 3.10+, ASCII-only output, handles Windows/Linux/macOS path differences.
Testing
# Test against a copy of the original binary
python3 patch.py --binary /path/to/claude_copy
Adding New Patches
When adding a patch, follow the existing pattern:
- Pick a stable anchor (string literal, property name — never a minified identifier)
- Use
find_backwards()/find_all()to locate the target byte - Verify the byte before replacing
- Expect 2x matches per anchor (or 4x if both plugin and server checks share the same anchor)
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.
- 5d ago First seen · 41 lines · 482 tokens per session scan A 9ef5c5d21620
claude-channels-patch CLAUDE.md is an instructions file published in the GitHub repository wuzf/claude-channels-patch (44 stars, last pushed 5mo ago), licensed MIT. It adds 482 tokens to every session, about $0.0024 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 2026-08-30.
Other instructions, from other repositories
cs CLAUDE.md
Claude Code instructions for boyter/cs, covering claude.md, project overview, build & test commands, architecture and query pipeline (pkg/search/).
gmgn-skills CLAUDE.md
Claude Code instructions for GMGNAI/gmgn-skills, covering claude.md, critical rule — read this first, project overview, available skills and quick decision guide.
hana-developer-cli-tool-example mcp-server-development.instructions.md
Use when creating or updating MCP (Model Context Protocol) server components. Enforces TypeScript patterns, JSON-RPC communication rules, tool/resource/prompt registration, and integration with CLI command metadata. Ensures MCP server follows protocol requirements and maintains consistency with the CLI infrastructure.
hana-developer-cli-tool-example route-development.instructions.md
Use when creating or updating route files in routes/. Enforces consistent structure, error handling, database connection management, internationalization, and integration with the Express server and CLI command infrastructure. Ensures routes follow established patterns for endpoint registration, response formatting…
hana-developer-cli-tool-example connection-pooling-timeout-management.instructions.md
Use when creating or updating code that manages database connections. Enforces consistent timeout configuration, proper connection cleanup, graceful error handling, and connection pooling best practices across CLI commands, routes, and tests.
core copilot-instructions.md
Copilot instructions for asdecided/core: These decisions are already accepted. Do not re-open or contradict them; ask the AsDecided MCP tools (getartifact, searchartifacts) for the full text before proposing a change that touches one.