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/vakovalskii/phantom-agent/claude-mdgit clone --depth 1 https://github.com/vakovalskii/phantom-agentWrote 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/vakovalskii/phantom-agent/claude-md)<a href="https://agentmods.dev/instructions/vakovalskii/phantom-agent/claude-md"><img src="https://agentmods.dev/badge/instructions/vakovalskii/phantom-agent/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 | $0.01976 | $0.01976 |
| Opus 5 | $0.00988 | $0.00988 |
| Sonnet 5 | $0.00395 | $0.00395 |
| Haiku 4.5 | $0.00198 | $0.00198 |
Grade A, and why
phantom-agent 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.
How it starts
The opening of the file, as written. The whole thing — 181 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PAC1 Benchmark Agent
Quick Start
uv sync
cd dashboard && npm install && cd ..
# Start backend
uv run python server.py
# Start frontend (in another terminal)
cd dashboard && npm run dev
Dashboard: http://localhost:5173 | API: http://localhost:8000
Credentials via .env file (not committed) or Settings tab in dashboard.
Restarting Services
Backend does NOT auto-reload. After changing Python files:
kill $(lsof -t -i :8000) && uv run python server.py
Frontend (Vite) hot-reloads automatically — no restart needed for JSX/CSS changes.
Architecture
v2 Agent — Pure LLM ReAct on OpenAI Agents SDK. Progressive skill disclosure.
User task → LLM Classifier (picks skill)
→ Agent receives: system_prompt (5.2K) + skills_menu + task + skill_hint
→ Step 1: get_skill_instructions(recommended_skill) — loads full workflow
→ ReAct loop: LLM → tool call → result → LLM → ... → submit_answer
→ Fallbacks: retry on text-only/ModelBehaviorError → force-tool agent → error recovery
Key Files
| File | Purpose |
|---|---|
main_v2.py |
CLI benchmark runner (sliding window parallelism) |
server.py |
FastAPI + SSE backend for dashboard |
agent_v2/agent.py |
Agent creation, run_task, retries, force-tool, Harmony patches |
agent_v2/prompts.py |
System prompt + skills menu builder + task prompt builder |
agent_v2/system_prompt.md |
System prompt: MAIN_ROLE, APPROACH, SECURITY, CONSTRAINTS, COMPLETION |
agent_v2/tools.py |
13 tools via @function_tool, auto-merge grounding_refs |
agent_v2/skills/ |
12 skill prompts (.md) + classifier + LLM classifier |
agent_v2/hooks.py |
Live logging hooks (console + SSE) |
agent_v2/runtime.py |
Async PCM gRPC wrapper |
agent_v2/db.py |
SQLite persistence (runs, tasks, events) |
agent_v2/config.py |
Env config via .env / os.getenv |
agent_v2/verifier.py |
Optional outcome verifier (kimi-k2.5, currently unused) |
Prompt Architecture (Progressive Disclosure)
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 · 181 lines · 1,976 tokens per session scan A d17b1ea8430f
phantom-agent CLAUDE.md is an instructions file published in the GitHub repository vakovalskii/phantom-agent (32 stars, last pushed 4mo ago), licensed MIT. It adds 1,976 tokens to every session, about $0.0099 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
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
deepseek-harness AGENTS.md
AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.