wa CLAUDE.md

wa CLAUDE.md is an instructions file for Claude Code from yolo-labz/wa. It costs 15,118 tokens per session, scanned B, original, Apache-2.0.

A project instruction file for the wa repository that records its current status, architecture decisions, reliability rules, and development guidance. It says the repository is an early Go project with its structure prepared but no Go source files yet.

In plain words
What is it for?
Use it when working on the wa codebase to understand its plans, policies, directory structure, and reliability expectations. It also explains which tracked documents take priority.
Why use it?
It gives coding agents a shared reference for decisions and project rules, reducing the chance that work contradicts the repository’s agreed design or status.

Instructions file for Claude Code

Written for Claude Code: ${CLAUDE_PLUGIN_ROOT} variable. Also seen: reads .claude/ paths; mentions CLAUDE.md; mentions Claude Code.

Runs only inside a plugin — its command needs a path that Claude Code sets for a plugin’s own hooks and for nothing else, and the catalogue could not identify which plugin ships it.

Install

Getting it into your agent

There is no command for this one: it runs only inside a plugin, and the catalogue could not identify which plugin ships it. The source is linked below.

Made for: Claude Code.

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.

agentmods badge for wa CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/yolo-labz/wa/claude-md.svg)](https://agentmods.dev/instructions/yolo-labz/wa/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/yolo-labz/wa/claude-md"><img src="https://agentmods.dev/badge/instructions/yolo-labz/wa/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 15,118 This file is loaded in full into every session.
When invoked 15,118 The same file — it is already loaded in full.
Security scan B 2 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 5.1 $0.15118 $0.15118
Opus 5 $0.07559 $0.07559
Sonnet 5 $0.03024 $0.03024
Haiku 4.5 $0.01512 $0.01512

Measured 5d ago against content hash ed28a6f2e48c, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade B, and why

wa CLAUDE.md scanned grade B with 2 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.

Asks for rootmediumPrivilege escalation

A mod that escalates privileges can change anything on the machine, not only the project.

- **Channel state lives at `~/.claude/channels/wa/`**, mirroring Telegram's layout: `access.json` (allowlist, pending pairings, dmPolicy) is hand-edited only by `/wa:access`; `.env` (any future tokens) is `chmod 0600`. T

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

- **Bootstrap** of the `wa`/`wad` binaries does NOT happen via a plugin install lifecycle hook — Claude Code plugins have no `scripts.postInstall` field (verified against the official Telegram plugin source 2026-04-06).
CLAUDE.md · 593 lines

How it starts

The opening of the file, as written. The whole thing — 593 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Project status

Pre-source. As of 2026-04-06 the repository contains the architectural blueprint (this file), the spec/plan/research/contracts/quickstart for features 001-research-bootstrap (closed) and 002-domain-and-ports (planning), the hexagonal directory skeleton with .gitkeep placeholders, the full governance file set (LICENSE, README, SECURITY, .gitignore, .editorconfig, .golangci.yml, cliff.toml, renovate.json, lefthook.yml, .github/workflows/ci.yml), and go.mod. Zero *.go source files exist yet — feature 002's /speckit:implement writes the first ones. Treat every section below as a decision already made unless explicitly flagged otherwise. The tracked policy in this file plus tracked amendment records under specs/ is the PR-reviewable constitution authority. .specify/memory/constitution.md is an ignored generated local mirror: regenerate it from tracked policy when useful, but never let it override a tracked rule or amendment. The reliability principles below are the high-attention summary of docs/reliability.md, which carries the citation trail.

Reliability principles (load-bearing)

These rules are placed near the top of CLAUDE.md deliberately. LLM attention degrades past line ~400 (Liu 2023, RULER 2024, NoLiMa 2025); rules buried in the middle of a long context measurably stop firing. The full citation trail and rationale live in docs/reliability.md, synthesised from a five-agent research swarm on 2026-04-06 (raw dossiers under docs/research-dossiers/).

Speckit workflow

  1. Constitution-first. Versioned, falsifiable principles MUST be settled before /speckit:specify. The review authority is tracked CLAUDE.md plus tracked amendment records under specs/; ignored .specify/memory/constitution.md is only a generated local mirror. Aspirational principles ("we value quality") are forbidden.
  2. Generated artefacts are regenerated, not hand-edited. spec.md, plan.md, tasks.md, research.md are produced by their slash commands. The "spec laundering" anti-pattern (agent edits the spec to match the code it just wrote) is forbidden.
  3. /speckit:clarify before /speckit:plan, always. No /plan may run with [NEEDS CLARIFICATION] markers in the spec.
  4. /speckit:analyze before /speckit:implement from feature 002 onward. Cross-artefact consistency check catches multi-feature drift.
  5. data-model.md is the single field authority. /implement may not reference any entity, field, or type that does not appear there. If it's missing, stop and re-run /speckit:plan.
  6. One feature in flight per branch; cap tasks.md at ~25 items. Split larger features.

Read the full file on GitHub · 593 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. 5d ago First seen · 593 lines · 15,118 tokens per session scan B ed28a6f2e48c

Subscribe to this mod's changes

wa CLAUDE.md is an instructions file published in the GitHub repository yolo-labz/wa (0 stars, last pushed 8d ago), licensed Apache-2.0. It adds 15,118 tokens to every session, about $0.0756 per session on Opus 5. A static security scan graded it B with 2 findings (asks for root, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.