contextforge

A project rule for using ContextForge, a tool that creates smaller, selected bundles of relevant project files for coding tasks. It also uses a dependency graph to show which files are connected.

In plain words
What is it for?
Use it to select task-specific code context, inspect file dependencies, create change specifications, and check a diff against project guardrails before committing.
Why use it?
It reduces the need to load an entire codebase into context and helps keep proposed changes focused on the files that matter.

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/alejandro-cedeno-10/contextforge-cli/contextforge
Clone the repo
git clone --depth 1 https://github.com/alejandro-cedeno-10/contextforge-cli

Made for: Cursor.

Per session 299 This file is loaded in full into every session.
When invoked 299 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 $0.00299 $0.00299
Opus 5 $0.00150 $0.00150
Sonnet 5 $0.00060 $0.00060
Haiku 4.5 $0.00030 $0.00030

Measured yesterday against content hash 716384b61fff, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

contextforge 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 yesterday.

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/contextforge.mdc · 38 lines

What it actually says

ContextForge integration

This project uses ContextForge — a deterministic pipeline that produces curated context artifacts to save tokens during AI-assisted development.

Before suggesting code changes

  1. Check if .contextforge/context-pack.json exists.
  2. If yes, those files are the primary context for the current task. Prefer them over reading the entire codebase.
  3. Use .contextforge/graph.json to understand which files depend on the ones being modified.

When artifacts are missing or stale

Suggest running this pipeline (deterministic, no LLM, < 30 s):

pnpm forge scan       # Index files with BLAKE3 hashes
pnpm forge graph      # Build dependency graph (tree-sitter)
pnpm forge context "<task description>"  # PageRank + budget pack

When generating specs

forge spec <change-id> always emits OpenSpec format at openspec/changes/<id>/. Reference these files when working on a tracked change.

When committing

Run pnpm forge implement --check before committing changes — validates the diff against guardrails (allowed files, max LOC delta, forbidden paths).

Token savings observed

Real measurement on this repo: 90.3 % savings vs full repo dump (11 700 vs 121 000 tokens).

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. yesterday First seen · 38 lines · 299 tokens per session scan A 716384b61fff

Subscribe to this mod's changes

contextforge is a cursor rule published in the GitHub repository alejandro-cedeno-10/contextforge-cli (0 stars, last pushed 3mo ago), licensed MIT. It adds 299 tokens to every session, about $0.0015 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.