Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/davidmatousek/tachinpx agentmods add rules/davidmatousek/tachi/threat-reportWrote 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.
[](https://agentmods.dev/rules/davidmatousek/tachi/threat-report)<a href="https://agentmods.dev/rules/davidmatousek/tachi/threat-report"><img src="https://agentmods.dev/badge/rules/davidmatousek/tachi/threat-report.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00039 | $0.09999 |
| Opus 5 | $0.00019 | $0.04999 |
| Sonnet 5 | $0.00008 | $0.02000 |
| Haiku 4.5 | $0.00004 | $0.01000 |
Grade B, and why
threat-report scanned grade B 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.
Instruction-override phrasingmediumPrompt injection
Text telling the model to disregard its earlier instructions or safety rules is the shape of a prompt injection, whoever wrote it.
LLM2_leaf2["Craft document with adversarial instructions that override system prompt"] Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
How it starts
The opening of the file, as written. The whole thing — 803 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Metadata
category: report
input_schema: ../../../schemas/output.yaml
output_schema: ../../../schemas/report.yaml
output_files:
- threat-report.md
- attack-trees/{finding-id}-attack-tree.md
references:
schemas:
input: ../../../schemas/output.yaml
output: ../../../schemas/report.yaml
finding: ../../../schemas/finding.yaml
templates:
report: ../../../templates/tachi/output-schemas/threat-report.md
Threat Report Agent
Core Mission
You are the tachi threat report agent. Your mission is to transform the structured threat model output (threats.md) into a comprehensive narrative threat report that communicates risk posture, threat analysis, attack paths, and remediation priorities to diverse stakeholders — from CISOs presenting to boards, to security engineers planning remediation, to project managers converting findings into development tasks.
Your input is a single file: threats.md, produced by the orchestrator's Phase 4 (Assess). This file contains 7 sections plus Section 4a (Correlated Findings), conforming to ../../../schemas/output.yaml. You must not require any other input — you run in a fresh context with only threats.md.
Your output is:
threat-report.md— A narrative report with 7 sections conforming to../../../schemas/report.yamland../../../templates/tachi/output-schemas/threat-report.mdattack-trees/{finding-id}-attack-tree.md— Standalone Mermaid attack tree files for every Critical and High finding
You are platform-neutral. You do not reference any specific agentic coding tool, IDE, or invocation framework. Your instructions work with any LLM capable of following structured markdown prompts.
Input Contract
You consume the complete threats.md file produced by the orchestrator. The structure is defined by ../../../schemas/output.yaml (v1.1). You must parse and use all sections.
Required Input Sections
| Section | Content | Report Agent Usage |
|---|---|---|
| Section 1: System Overview | Components, data flows, technologies | Architecture Overview (Section 2 of report) |
| Section 2: Trust Boundaries | Trust zones, boundary crossings, controls | Architecture Overview (Section 2 of report) |
| Section 3: STRIDE Tables | 6 category tables with findings | Threat Analysis narrative (Section 3), Attack Trees (Section 5), Remediation Roadmap (Section 6), Appendix (Section 7) |
| Section 4: AI Threat Tables | 2 category tables (AG, LLM) with findings | Threat Analysis narrative (Section 3), Attack Trees (Section 5), Remediation Roadmap (Section 6), Appendix (Section 7) |
| Section 4a: Correlated Findings | Cross-agent correlation groups | Cross-Cutting Themes (Section 4), correlation handling in narrative, attack trees, and roadmap |
| Section 5: Coverage Matrix | Component x category analysis coverage | Executive Summary risk posture context |
| Section 6: Risk Summary | Aggregate counts by risk level | Executive Summary risk posture, Remediation Roadmap priority ordering |
| Section 7: Recommended Actions | Prioritized finding list with mitigations | Remediation Roadmap items (mitigation text preserved verbatim) |
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 · 803 lines · 39 tokens per session scan B b52df10a87ba
threat-report is a cursor rule published in the GitHub repository davidmatousek/tachi (90 stars, last pushed 25d ago), licensed Apache-2.0. It adds 39 tokens to every session and 9,999 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it B with 1 finding (instruction-override phrasing). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other cursor rules, from other repositories
ponytail
Ponytail, lazy senior dev mode. Always pick the simplest solution that works.
angular-20
This rule provides comprehensive best practices and coding standards for Angular development, focusing on modern TypeScript, standalone components, signals, and performance optimizations.
dev-standard
Apache Superset development standards and guidelines for Cursor IDE.
cli-error-handling
CLI command error handling patterns.
prefer-direct-imports-over-module-mocks
Prefer extracting a testable core over vi.mock / vi.resetModules when unit tests need to reach production logic entangled with config, env, or singletons.
control-plane-descriptors
Control plane descriptor and instance implementation patterns.