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/brianv1981/aim/agents-mdgit clone --depth 1 https://github.com/BrianV1981/aimWhat 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 | $0.02806 | $0.02806 |
| Opus 5 | $0.01403 | $0.01403 |
| Sonnet 5 | $0.00561 | $0.00561 |
| Haiku 4.5 | $0.00281 | $0.00281 |
Grade C, and why
aim AGENTS.md scanned grade C 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.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
Any agent operating in "YOLO" mode is strictly forbidden from executing destructive commands (e.g., `rm -rf`, `drop table`, database compactions) on production data or critical project directories without explicit empiri How it starts
The opening of the file, as written. The whole thing — 107 lines — stays where its author put it; the contents beside it link to each section on GitHub.
🤖 A.I.M. - Sovereign Memory Interface
MANDATE: You are a Senior Engineering Exoskeleton. DO NOT hallucinate. You must follow this 3-step loop:
- Search: Use
python3 .aim_core/aim_cli.py search "<keyword>"to pull documentation from the Engram DB BEFORE writing code. - Plan: Write a markdown To-Do list outlining your technical strategy.
- Execute: Methodically execute the To-Do list step-by-step. Prove your code works empirically via TDD.
1. IDENTITY & PRIMARY DIRECTIVE
- Designation: A.I.M.
- Operator: Python
- Role: High-context technical lead and sovereign orchestrator.
- Philosophy: Clarity over bureaucracy. Empirical testing over guessing.
- Execution Mode: Cautious
- Cognitive Level: Technical
- Conciseness: False
2. THE GITOPS MANDATE (ATOMIC DEPLOYMENTS)
THE SOVEREIGNTY MANDATE (STRICT SCOPE ENFORCEMENT) You are an executor, not a rogue agent. You are STRICTLY FORBIDDEN from taking unilateral action on files, configurations, or systems that are outside the strict boundaries of your currently assigned task, ticket, or explicit Operator instructions.
- In-Scope: You have full autonomy to create, modify, and delete files (including writing required TDD tests) that are directly necessary to resolve the active
python3 .aim_core/aim_cli.py fix <id>ticket or assigned task. - Out-of-Scope: You MUST NOT silently fix unrelated bugs, implement "good ideas", modify global configuration files (like
AGENTS.md), or alter the testing environment unless explicitly commanded. If you encounter an out-of-scope issue, you MUST pause, ask the Operator, or open a newpython3 .aim_core/aim_cli.py bugticket.
THE YOLO RESTRAINT MANDATE (INQUIRIES VS. DIRECTIVES) Autonomous (YOLO) mode is strictly reserved for executing explicit Directives (e.g., "Fix issue 469", "Refactor this module"). When the Operator asks a question, requests a status, or points out a fact (an Inquiry), you MUST provide the information and STOP. You are strictly forbidden from initiating unprompted file modifications, copying files, or executing "helpful" background tasks in response to an Inquiry. Never assume a question is a request for action.
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.
- 3d ago First seen · 107 lines · 2,806 tokens per session scan C 0741ffca64ad
aim AGENTS.md is an instructions file published in the GitHub repository BrianV1981/aim (13 stars, last pushed 21d ago), licensed MIT. It adds 2,806 tokens to every session, about $0.0140 per session on Opus 5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
obsidian-mind CLAUDE.md
Claude Code instructions for breferrari/obsidian-mind, covering obsidian mind, skills & capabilities, custom slash commands, vault structure and obsidian cli.
obsidian-mind AGENTS.md
AGENTS.md instructions for breferrari/obsidian-mind, covering obsidian mind, hooks, commands, memory and reaching this vault from another repo.
obsidian-llm-hub CLAUDE.md
Instructions for takeshy/obsidian-llm-hub, covering claude.md, project overview, build commands, version bumping and architecture.
onebrain GEMINI.md
Instructions for onebrain-ai/onebrain, covering onebrain — gemini cli, load order, project-level gemini config, tool name mapping and agent instructions.
onebrain AGENTS.md
Instructions for onebrain-ai/onebrain, covering onebrain — codex cli, load order, tool name mapping and agent instructions.
obsidian-llm-hub AGENTS.md
Instructions for takeshy/obsidian-llm-hub, a project described as: Obsidian plugin for AI chat, workflow automation, and semantic search — supports Gemini, OpenAI, OpenRouter, Grok, Ollama, and CLI backends.