A set of rules for running a research-and-testing loop in repositories that contain specific Forge files. It uses repeated evaluations, project notes, and an Obsidian vault, a local knowledge base.

Use it to validate research documents, run the repository's evaluation script three times, compare scores, propose three hypotheses, and record results.

It gives experimental development a fixed startup, testing, reporting, and decision process.

forge — Cursor rule for Cursor by yash-dev007/THE-FORGE

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.

Clone the repo

git clone --depth 1 https://github.com/yash-dev007/THE-FORGE

Made for: Cursor.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/rules/yash-dev007/the-forge/forge/github.svg)](https://agentmods.dev/rules/yash-dev007/the-forge/forge)

Your own site

<a href="https://agentmods.dev/rules/yash-dev007/the-forge/forge"><img src="https://agentmods.dev/badge/rules/yash-dev007/the-forge/forge/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/rules/yash-dev007/the-forge/forge"><img src="https://agentmods.dev/badge/rules/yash-dev007/the-forge/forge.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 0 Nothing until a file matches its globs; then the whole rule loads.

When invoked 513 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

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.

Model	Per session	Once invoked
Fable 5.1	$0.00000	$0.00513
Opus 5.5	$0.00000	$0.00205
Sonnet 5.5	$0.00000	$0.00103
Haiku 4.5	$0.00000	$0.00051

Measured 1mo ago against content hash 5257918c7f53, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

forge 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 1mo 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.

templates/cursor/forge.mdc · 32 lines

What it actually says

THE FORGE v4 (multi-project)

When working in a repository that uses THE FORGE (presence of RESEARCH.md + EVAL_SPEC.md + EVAL.sh at the repository root), treat the project as a research loop.

Startup (each session)

Read FORGE_IDENTITY.md — extract ForgeProjectSlug and ObsidianVaultRoot. If missing → [FORGE] BLOCKED.
Read the Quintet in order: RESEARCH.md → EVAL_SPEC.md → PROJECT_LOG.md → FORGE_SYSTEM.md → CLAUDE.md.
Validate RESEARCH.md contains all eight required sections. If any blank/placeholder → [FORGE] BLOCKED.
Run the loop driver: bash ./forge-cycle.sh --baseline-only — runs EVAL × 3, computes median, prints Status Report.
Read Obsidian patterns under Forge/Patterns/ and Forge/Projects/<ForgeProjectSlug>/.
Propose exactly three hypotheses ranked by P(success). Wait for human selection.

During work

Start every response with the [FORGE] status line: [FORGE] Cycle N | Type: X | Hypothesis: "..." | Score: X.X → Y.Y | Δ: ±Z.Z | Decision: ...
After human implements: run bash ./forge-cycle.sh --skip-baseline <SCORE> — script handles eval, decision, log write, Obsidian sync.
Do not modify EVAL.sh or EVAL_SPEC.md during hypothesis cycles (maintainer-only).
On ANOMALY output → [FORGE] ANOMALY DETECTED and guide revert.
Auditor every five commits; Pattern Distillation every ten cycles — counts in this repo's PROJECT_LOG.md only.

If Quintet missing

Do not invent scores. Point the user to THE FORGE kit docs/ADOPT.md (path may be a sibling repo or where they installed the kit).

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.

1mo ago First seen · 32 lines · 0 tokens per session scan A 5257918c7f53

Subscribe to this mod's changes

forge is a cursor rule published in the GitHub repository yash-dev007/THE-FORGE (2 stars, last pushed 6mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 513 tokens. 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.

Research Testing

forge

Getting it into your agent

Wrote this? Show the measurements

What it costs to keep this loaded

Grade A, and why

Nothing flagged

What it actually says

THE FORGE v4 (multi-project)

Startup (each session)

During work

If Quintet missing

What this file has done since we first saw it

Other cursor rules, from other repositories

20-testing-generated-and-structure

fuzzing

testing-and-quality

smoke-monkey-tester

testing

STAGE_S5_TEST_POINTS