almanac CLAUDE.md

almanac CLAUDE.md is an instructions file for coding agents from corcoran/almanac. It costs 6,316 tokens per session, scanned A, original, BSD-2-Clause.

A set of project instructions for Claude Code covering Almanac, including local notes, technical documentation, code navigation, and writing style. Claude Code is an AI coding assistant.

In plain words
What is it for?
It helps guide code lookup, references, comments, changelogs, documentation, and user-facing writing in the Almanac project.
Why use it?
It gives the assistant repository-specific rules, including how to handle private planning files and navigate a pnpm monorepo, a repository containing multiple related packages.

Instructions file

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 instructions/corcoran/almanac/claude-md
Clone the repo
git clone --depth 1 https://github.com/corcoran/almanac

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 almanac CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/corcoran/almanac/claude-md.svg)](https://agentmods.dev/instructions/corcoran/almanac/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/corcoran/almanac/claude-md"><img src="https://agentmods.dev/badge/instructions/corcoran/almanac/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 6,316 This file is loaded in full into every session.
When invoked 6,316 The same file — it is already loaded in full.
Security scan A 1 finding. 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 $0.06316 $0.06316
Opus 5 $0.03158 $0.03158
Sonnet 5 $0.01263 $0.01263
Haiku 4.5 $0.00632 $0.00632

Measured 3d ago against content hash 868d31bef2f3, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

almanac CLAUDE.md scanned grade A with 1 finding 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 3d 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.

Makes network callslowCapability

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

injects it for browser requests (so the UI needs no login); for raw curl, pass the
CLAUDE.md · 441 lines

How it starts

The opening of the file, as written. The whole thing — 441 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Almanac — Agent Notes

Specs and plans are local-only

docs/superpowers/ (specs + plans) is gitignored (.gitignore:14) — working notes, deliberately not published with the repo. So they have no git history and no diff to recover from: treat an edit there as destructive, and don't reach for git diff/git status to review changes to one.

Code navigation (LSP)

A TypeScript LSP tool is available (the typescript-lsp plugin, backed by typescript-language-server). Prefer it over grep for symbol-level queries — it's type-resolved, so it ignores dist/ build output, comments, and same-name false matches. Use it for: go-to-definition, find-references, hover (inferred types), document/workspace symbols, and call hierarchy (who-calls-X / X-calls).

Two caveats specific to this repo:

  • Aim the cursor inside the identifier. Positions are 1-based; targeting the very start of a declaration can resolve to the keyword instead of the name. Read the line first, then point a few chars into the symbol.
  • It's package-scoped — it misses cross-package barrel imports. This is a pnpm monorepo where api/web/mcp import core symbols via @almanac/core/* barrels (not direct source paths), so findReferences on a core symbol won't list its api/web call sites. For an exhaustive monorepo-wide sweep, pair LSP with a grep on the imported name (as the release-removal work needed).

Prose: comments, changelog, docs

One operator, one deployment — nobody else has installed this. Write for someone who wants to get on with the task, not for a stressed stranger migrating a fleet. Verbosity is the default failure mode here — every pass has needed cutting, never expanding.

There is no installed base, so there are no upgrade notes. Never write "on an existing deployment…", "before updating, add…", or a migration path for a change that shipped in the same release. If a change needs an action, the user takes it once, at deploy time, from the release notes — not from a permanent warning box aimed at operators who do not exist. Document the current state of things; the fact that it used to be different is git history.

Read the full file on GitHub · 441 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. 3d ago First seen · 441 lines · 6,316 tokens per session scan A 868d31bef2f3

Subscribe to this mod's changes

almanac CLAUDE.md is an instructions file published in the GitHub repository corcoran/almanac (1 stars, last pushed 18d ago), licensed BSD-2-Clause. It adds 6,316 tokens to every session, about $0.0316 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.