supervisor-specs

Cursor rule "supervisor-specs" from ai-supervisor-foundry/foundry, covering supervisor specifications and anti-goals (do not implement).

Cursor rule for Cursor

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 rules/ai-supervisor-foundry/foundry/supervisor-specs
Clone the repo
git clone --depth 1 https://github.com/ai-supervisor-foundry/foundry

Made for: Cursor.

Per session 167 This file is loaded in full into every session.
When invoked 167 The same file — it is already loaded in full.
Security scan A 0 findings. 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 $0.00167 $0.00167
Opus 5 $0.00084 $0.00084
Sonnet 5 $0.00033 $0.00033
Haiku 4.5 $0.00017 $0.00017

Measured today against content hash b1f2d15e81db, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

supervisor-specs 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 today.

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.

.cursor/rules/supervisor-specs.mdc · 31 lines

What it actually says

Supervisor Specifications

  • The supervisor does not define goals.
  • Operator must inject goals.
  • Scope cannot be expanded by AI.
  • All tasks require explicit acceptance criteria.
  • No task runs without validation.
  • No refactoring without explicit instruction.
  • State must be persisted after every step.
  • Ambiguity halts execution.
  • Cursor CLI is a worker tool, not decision authority.
  • AUTO MODE is default and mandatory.
  • AUTO MODE cannot be disabled without operator instruction.
  • No silent retries.
  • All outputs are reviewable (diffs, logs).

Anti-Goals (Do NOT Implement)

  • Autonomous goal refinement
  • Speculative task creation
  • Retry heuristics
  • AI-based validation
  • "Helpful" corrections
  • Fallback behaviors

If tempted → HALT.

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. today First seen · 31 lines · 167 tokens per session scan A b1f2d15e81db

Subscribe to this mod's changes

supervisor-specs is a cursor rule published in the GitHub repository ai-supervisor-foundry/foundry (11 stars, last pushed 1mo ago), licensed MIT. It adds 167 tokens to every session, about $0.0008 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 2026-09-02.