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/nirvanaguha/thoth/claude-mdgit clone --depth 1 https://github.com/NirvanaGuha/thothWhat 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 | $0.01604 | $0.01604 |
| Opus 5 | $0.00802 | $0.00802 |
| Sonnet 5 | $0.00321 | $0.00321 |
| Haiku 4.5 | $0.00160 | $0.00160 |
Grade D, and why
thoth CLAUDE.md scanned grade D with 3 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 2d 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.
Enumerates other installed skillsmediumAgent snooping
Other skills' SKILL.md files reveal prompts, capabilities and secrets that should be invisible to peers.
ls .claude/skills/thoth Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
rm -rf .claude Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
├── install.sh ← curl|bash installer (no Node required) How it starts
The opening of the file, as written. The whole thing — 115 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Thoth — Developer Guide
This file is for Claude (or a human dev) working on the Thoth skill itself, not for end-users. Read this alongside CONTRIBUTING.md.
Repo anatomy
thoth/
├── skill/ ← the skill. SOURCE OF TRUTH.
├── cli/ ← npm package; wraps installer logic around skill/
│ ├── bin/thoth.js ← zero-dep Node CLI
│ └── assets/skill/ ← auto-synced copy for --offline installs
├── .claude-plugin/ ← Claude plugin marketplace manifest
├── install.sh ← curl|bash installer (no Node required)
├── skill.json ← skill-level metadata
└── .github/workflows/ ← CI that publishes the CLI on release
The golden rule: all skill content lives in skill/. Never edit cli/assets/skill/ directly — it's re-synced from skill/ at release time. If you edit both, skill/ wins.
Code vs data separation: the skill/ directory is immutable skill code. Persona data does not live here. At runtime Thoth resolves a data root — ./.thoth/ if present in CWD, else ~/.thoth/ (default), else legacy ~/.claude/skills/thoth/ for pre-v1.1 installs. The skill/personas/ directory in this repo is intentionally empty except for README.md (which documents the resolution rules). The resolution algorithm and one-time migration logic are defined in skill/SKILL.md under "Where persona data lives" — do not bypass it by hardcoding paths anywhere else in the skill.
Conventions
- Ratio numbers appear in several files (
SKILL.md,skill/references/content-mix.md,skill/references/post-types.md,README.md). When you change one, grep for the old number and change them all. Ratios must always sum to 100. - Field names in persona.md are snake_case (
anti_voice,contrarian_beliefs,signature_grievances,archetype.dominant,tone.formal_casual). Prose can say "anti-voice" as a concept. YAML references in backticks must use snake_case. - Archetype names are canonical (Hero, Outlaw, Magician, Everyman, Lover, Jester, Caregiver, Creator, Ruler, Innocent, Sage, Explorer). "Regular Guy" is a Pearson/Mark alias for Everyman — OK to mention in parentheses inside
brand-archetypes.md, never as a primary name elsewhere. - Commands — any command added must be registered in:
skill/SKILL.mdcommand tableskill/references/commands.mdREADME.mdcommand table- SKILL.md dispatch section (with its logic described)
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.
- 2d ago First seen · 115 lines · 1,604 tokens per session scan D 9f8612f93a32
thoth CLAUDE.md is an instructions file published in the GitHub repository NirvanaGuha/thoth (3 stars, last pushed 2mo ago), licensed MIT. It adds 1,604 tokens to every session, about $0.0080 per session on Opus 5. A static security scan graded it D with 3 findings (enumerates other installed skills, recursive force delete, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
pm-skills AGENTS.md
Instructions for product-on-purpose/pm-skills, covering pm-skills, skills, foundation classification, discover phase and define phase.
pm-skills CLAUDE.md
Instructions for product-on-purpose/pm-skills, covering claude code instructions, style rules, no em-dash characters, documentation rules and public vs private files.
llm-context.py CLAUDE.md
Instructions for cyberchitta/llm-context.py, covering claude.md, working notes (gitignored) and draining the field notes.
designing-real-world-ai-agents-workshop CLAUDE.md
Instructions for iusztinpaul/designing-real-world-ai-agents-workshop, covering project, project structure, tech stack, access documentation and running qa.
a-i--skills 03-development-workflow.instructions.md
Instructions for organvm-iv-taxis/a-i--skills, covering development workflow instructions, no repo-wide build system, common development commands, after adding/modifying skills and validation before committing.
rag-code-mcp copilot-instructions.md
Instructions for doITmagic/rag-code-mcp, covering copilot instructions - ragcode mcp, ⚖️ the golden rule, project overview, architecture & patterns and developer workflows.