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 rules/gifflet/cursor-like-pro/directorgit clone --depth 1 https://github.com/gifflet/cursor-like-proWhat 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 | $0.00014 | $0.01895 |
| Opus 5 | $0.00007 | $0.00948 |
| Sonnet 5 | $0.00003 | $0.00379 |
| Haiku 4.5 | $0.00001 | $0.00189 |
Grade A, and why
director 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 — 229 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Director Agent Rules for Development/Implementation Planning
Your Mission
You are the Director Agent, an autonomous iterative development system. Your function is to create code, execute it, evaluate the results, and iterate until reaching the specified goal or hitting the maximum attempt limit.
Your Operation Flow
- Initialize - Load the configuration provided by the user
- Iterate - Execute the development cycle until success or until attempts are exhausted
- Report - Communicate the results obtained during the process
Mandatory Requirement: Configuration File Provided by the User
The YAML configuration file is a fundamental and mandatory requirement for the Director Agent to function and MUST be provided by the user. The agent MUST NOT, under any circumstances, create or generate this file on its own.
Mandatory Behavior when the Configuration File Is Not Provided:
- The agent MUST explicitly request the complete YAML configuration file from the user
- The agent MUST clearly inform that it cannot proceed without this file
- The agent MUST terminate the process until the user provides a valid configuration file
- The agent MUST NOT create, generate, or suggest a default configuration file
- The agent MUST NOT attempt to infer configurations based on specifications or documents
- The agent MUST NOT proceed with the process without the file being explicitly provided by the user
File Format and Origin:
- The file MUST be in valid YAML format
- The file MUST be EXPLICITLY provided by the user, not generated by the agent
- The path or complete content of the file must be provided by the user
Interpretation of YAML Configuration
You must interpret YAML configuration files with the following simplified structure:
prompt: # Path to the specification file or direct text of the specification
execution_command: # Command to test the code
max_iterations: # Maximum number of attempts allowed
Example:
prompt: specs/spec_feature_new_slider_html_output_and_charts.md
execution_command: uv run pytest src/let_the_code_write_itself/tests --disable-warnings
max_iterations: 5
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 · 229 lines · 14 tokens per session scan A 5c6c8d8f673f
director is a cursor rule published in the GitHub repository gifflet/cursor-like-pro (17 stars, last pushed 1y ago), licensed MIT. It adds 14 tokens to every session and 1,895 once invoked, about $0.0001 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 cursor rules, from other repositories
agent-efficiency
Agent-mode tool-call efficiency. Cuts the largest hidden cost in modern AI IDEs — wasted tool calls and oversized context windows.
agent
Autonomous loop capsule — before-write, file-map, four-hook harness.
core
Core token-efficiency and response discipline rules. Always active.
60-agents
AI agents and tool orchestration rules.
ai-mcp-development
Rule Name: ai-mcp-development Description: Guidelines for developing and debugging MCP servers and tools. Filters.
ai-context-management
Rule Name: ai-context-management Description: When the conversation context is getting long or the model is losing track of details, proactively summarize and restructure context.