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/asimons81/hardproof/agents-mdgit clone --depth 1 https://github.com/asimons81/hardproofWhat 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 | $0.03105 | $0.03105 |
| Opus 5 | $0.01553 | $0.01553 |
| Sonnet 5 | $0.00621 | $0.00621 |
| Haiku 4.5 | $0.00311 | $0.00311 |
Grade A, and why
hardproof AGENTS.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 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.
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.
How it starts
The opening of the file, as written. The whole thing — 297 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Hardproof Agent Instructions
This file loads as project context when Hermes Agent starts in the Hardproof repository root. Read it before making any changes.
Project Identity
- Hardproof is a persistent, risk-aware engineering protocol for Hermes Agent.
- Current public release: v1.0.1 Proven (PyPI, GitHub).
- Current development boundary: none scheduled; the v0.4.0–v1.0.1 arc is released.
- Hardproof is a standalone Hermes plugin discovered through the
hermes_agent.pluginsentry-point group. - It uses only public Hermes registration, hook, command, skill, and dispatch APIs.
- The plugin remains opt-in. Hardproof does not modify Hermes core files.
Read Order
Inspect these files in order before changing code:
AGENTS.md— this fileREADME.md— project overview and quick startSTATUS.md— current release and maintenance stateROADMAP.md— roadmap and version planCONTRIBUTING.md— development workflowdocs/architecture.md— system architecturedocs/profiles.md— run profiles (Quick, Standard, Critical)docs/security-model.md— security boundarydocs/configuration-and-migrations.md— config and migration guidedocs/policy-packs.md— language policy packs- Relevant ADRs in
docs/adr/ - Relevant tests in
tests/
Do not assume the README is the entire specification.
Architecture Map
hardproof/
├── __init__.py # Package version and public API
├── plugin.py # Plugin registration entry point
├── compat.py # Hermes API compatibility checks
├── config.py # Configuration loading and defaults
├── constants.py # Shared constants
├── errors.py # Error types
├── paths.py # Project-local path resolution
├── commands/
│ ├── cli.py # hermes hardproof CLI adapter
│ ├── slash.py # /hardproof slash-command adapter
│ └── shared.py # Shared command implementation
├── domain/
│ ├── enums.py # RunStage, RunProfile, ApprovalGate, etc.
│ ├── models.py # Run, VerificationCheck, SessionBinding, etc.
│ ├── snapshots.py # Workspace state snapshots
│ ├── transitions.py # Stage transition logic
│ └── workcells.py # Workcell task/attempt state machines
├── hooks/
│ ├── context.py # Pre-turn context injection
│ ├── sessions.py # Session lifecycle hooks
│ ├── tool_policy.py # Tool-call policy enforcement
│ └── verification.py # Verification hook
├── isolation/
│ ├── guard.py # Workspace isolation guard
│ └── adapters.py # Execution backend adapters
├── kanban/
│ ├── worker.py # Workcell implementer subagent contract
│ └── approvals.py # Workcell approval routing
├── migrations/ # SQL migration files
├── policy/
│ ├── packs.py # Language policy packs (Python, Node, Rust, Go)
│ ├── profiles.py # Profile definitions
│ ├── stage_graph.py # Stage graph configuration
│ ├── stage_rules.py # Stage-aware rules
│ ├── terminal.py # Terminal normalization
│ ├── tool_rules.py # Tool-call rules
│ ├── trace.py # Policy trace model
│ ├── verification_rules.py# Verification-specific rules
│ └── waivers.py # Waiver logic
├── services/
│ ├── approvals.py # Human approval service
│ ├── artifacts.py # Run artifact service
│ ├── authority.py # Human authority registry
│ ├── decisions.py # Policy decision persistence
│ ├── evidence.py # Verification evidence service
│ ├── fix_re_review.py # Challenge Chamber fix/re-review loop
│ ├── hermes_children.py # Public Hermes child-session adapter
│ ├── packages.py # Package metadata services
│ ├── reports.py # Report generation
│ ├── review_dispatch.py # Challenge Chamber reviewer dispatch
│ ├── risks.py # Risk classification
│ ├── runs.py # Run management
│ ├── sessions.py # Session binding
│ ├── tasks.py # Task management
│ ├── waivers.py # Waiver service
│ ├── workcell_artifacts.py# Workcell artifact handling
│ └── workcells.py # Workcell scheduling and claims
├── skills/ # 9 stage skill modules
│ ├── orchestrate/
│ ├── discover/
│ ├── design/
│ ├── plan/
│ ├── implement/
│ ├── review/
│ ├── verify/
│ ├── deliver/
│ └── learn/
├── storage/
│ ├── database.py # SQLite database wrapper
│ ├── migrations.py # Migration engine
│ └── repository.py # Run repository
├── templates/ # Report templates
│ ├── completion.md
│ ├── design.md
│ ├── discovery.md
│ ├── plan.md
│ └── review.md
├── validation/ # Protocol SDK validation interfaces
├── vault/
│ └── lease.py # Vault lease state
└── tools/
├── handlers.py # Tool handler implementations
└── schemas.py # Exact JSON schemas for 6 tools
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 · 297 lines · 3,105 tokens per session scan A d0f09b4f737f
hardproof AGENTS.md is an instructions file published in the GitHub repository asimons81/hardproof (5 stars, last pushed 3d ago), licensed Apache-2.0. It adds 3,105 tokens to every session, about $0.0155 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 2026-08-31.
Other instructions, from other repositories
dograh AGENTS.md
AGENTS.md instructions for dograh-hq/dograh, covering dograh - project overview, project structure, tech stack, local development and environment configuration.
Patter AGENTS.md
Instructions for PatterAI/Patter, covering agents.md, non-negotiables (ci and reviewers block on these), before opening a pr, where things live and commits & prs.
windmill AGENTS.md
AGENTS.md instructions for windmill-labs/windmill, covering windmill, workflow, documentation, dev environment and per-worktree ports and database.
mainbranch AGENTS.md
Instructions for noontide-co/mainbranch, covering agent instructions, product shape, daily operating loop, quick start and repository layout.
basic-memory CLAUDE.md
Claude Code instructions for basicmachines-co/basic-memory, a project described as: AI conversations that actually remember. Never re-explain your project to your AI again. Join our Discord: https://discord.gg/tyvKNccgqN.
wet-mcp AGENTS.md
Instructions for n24q02m/wet-mcp, covering agents.md - wet-mcp, cau truc, lenh thuong dung, mise shortcuts and cau hinh quan trong.