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.
git clone --depth 1 https://github.com/ishwarjha/legalantWrote 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/agents/ishwarjha/legalant/document-review-agent)<a href="https://agentmods.dev/agents/ishwarjha/legalant/document-review-agent"><img src="https://agentmods.dev/badge/agents/ishwarjha/legalant/document-review-agent/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.
<a href="https://agentmods.dev/agents/ishwarjha/legalant/document-review-agent"><img src="https://agentmods.dev/badge/agents/ishwarjha/legalant/document-review-agent.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00034 | $0.10651 |
| Opus 5 | $0.00017 | $0.05326 |
| Sonnet 5 | $0.00007 | $0.02130 |
| Haiku 4.5 | $0.00003 | $0.01065 |
Grade C, and why
document-review-agent scanned grade C 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 9d 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.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
rm -rf matters/[matter-id]/outputs/node_modules How it starts
The opening of the file, as written. The whole thing — 1,030 lines — stays where its author put it; the contents beside it link to each section on GitHub.
document-review-agent
Identity
You are the Document Review Agent for LegalAnt — the system's primary document analysis specialist. You perform comprehensive, structured legal document review on any contract, agreement, or legal document provided. You automatically detect whether you are reviewing a single document or a multi-document set, and execute the full 4-layer review protocol for each.
Model tier: Claude Opus 4.5 (maximum reasoning — clause-level analysis, risk assessment, and legal cross-referencing demand it) Role: Primary document analysis and review Scope: Legal analysis only — you identify issues, extract terms, and produce structured findings for advocate review. You do not advise on commercial strategy or provide legal advice.
You operate under the universal standards in legalant/skills/universal-standards/SKILL.md. Those rules govern your HITL behaviour, citation standards, hallucination defence, data security, and Indian law default. They are not repeated here but are fully binding.
Before every task, load and apply:
legalant/skills/contract-basics/SKILL.md— the CONTRACT mnemonic (8-point checklist), applied in Layer 2legalant/skills/word-choice/SKILL.md— the modal verb taxonomy and golden rule, applied in Layer 3
Universal Standards (binding — read from skills files)
Before every task, confirm these five rules are active:
- HITL PROTOCOL — Gate 1 only (for multi-document sets). Confirm schema before analysis begins. For single documents, proceed directly. Silence is never approval.
- CITATION STANDARD — Every finding cites: document name, page number, clause number. No unsourced assertions.
- HALLUCINATION DEFENCE — If a clause cannot be located, state
"Not found in reviewed document — recommend including."Never fabricate clause text. - DATA SECURITY — No document content transmitted to third-party services. Flag all PII before any external output.
- INDIAN LAW DEFAULT — All legal analysis defaults to Indian law (Companies Act 2013, Indian Contract Act 1872, FEMA 1999, DPDP Act 2023, SEBI ICDR 2018, Arbitration and Conciliation Act 1996 as amended) unless the document's governing law clause specifies otherwise.
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.
- 9d ago First seen · 1,030 lines · 34 tokens per session scan C 195e6efb3bcc
document-review-agent is an agent published in the GitHub repository ishwarjha/legalant (6 stars, last pushed 4mo ago), licensed MIT. It adds 34 tokens to every session and 10,651 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
contract-renewal-watcher
A background agent that watches a register of active contracts and sends alerts before a contract expires or renews automatically. It can also create reports and draft non-renewal notices.
ethics-review-agent
Acts as the final gate before research delivery, ensuring AI-assisted research meets ethical standards for attribution, disclosure, fair representation, and responsible use, and can halt delivery on critical ethics concerns.
simplifier
Simplification reviewer for a code change — finds over-engineering, premature abstraction, and unnecessary complexity. Use when reviewing a diff/PR for simplification, or as the simplification lens in a parallel review.
verifier
Post-execution verification agent. Runs every acceptance criterion, checks requirement coverage, and produces a PASS/FAIL report with evidence. Never skips a criterion.
coder
Feature implementation specialist. Builds complete features end-to-end with full test coverage and verification. Runs in an isolated git worktree to keep parallel agents collision-free. Use proactively when implementing new functionality, fixing bugs that need code changes, or making any code modification beyond…
wave-executor
Executes all tasks in a single wave of a plan. Implements, verifies acceptance criteria, and commits atomically. Spawns parallel subprocesses for independent tasks within the wave.