REA

# Fix list: REA

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/rea, the October 2026 research run, assessed 12 October 2026. Grade B, 69.7 out of 100.

This is everything the published grade says the listing lacks, the biggest possible gain to the total first. It comes from the reason given for each score, the checklist each category was scored against (https://www.anchorterminal.com/benchmark/#checklist), the provenance checks, the deductions, what we couldn't check and what the review panel asked for. A fix counts at the next check, once it's public.

For a coding agent working on REA: work through the items below in the product, its docs and its public pages. Each category gives the reason for its score, with the points each checklist item earned, and the checklist itself, so the gap is the items that earned less than their points. Change the product, not the wording, and keep a note of what you changed and where it's published.

## 1. Security & auth, 49 out of 100, up to 8.9 more on the total

Why it scored 49: Local software with no credentials to leak. Provider bridges use a random capability token over a current-user Unix socket, so the middle band (20). Annotations mark destructive tools, and Windows Ghidra is read-only, but docs/mcp-contracts.md says REA requires no permission grants or per-call approval. Process, Electron, browser and adb install or uninstall tools run once the client allows them (6). Strings, bundles and pages from the target reach the model. Only WebMCP declarations are labelled untrusted, and cookies and authorisation headers are dropped from captures. No prompt-injection guidance (4). Every call is retained as an Evidence record in an exportable session bundle (10). SECURITY.md with private reporting and a written security boundary, npm trusted publishing with SLSA provenance. No security.txt, bug bounty or published advisories (9). Total 49.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-security):

- 0 to 30, the credential model. 30 for OAuth 2.1 with scopes, or scoped and revocable keys with rotation. 20 for plain revocable API keys. 10 for one all-powerful key. 10 off when a secret can travel in a URL query string as a documented option.
- 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions.
- 0 to 15, prompt-injection posture where the tool returns untrusted content (documented mitigations or guidance). A tool that returns no untrusted content gets 10.
- 0 to 15, audit logs or per-call visibility for the operator.
- 0 to 20, a security programme. security.txt or a disclosure policy, a bug bounty, SOC 2 or ISO 27001, advisories handled in public.

Models are read for retention, whether API data trains models (and whether that's off by default), zero-retention options and certifications. Frameworks for telemetry defaults, approval hooks, guardrails and sandboxing.

## 2. Payments & pricing, 60 out of 100, up to 5 more on the total

Why it scored 60: Free, self-hosted, MIT, nothing to buy from the project, so 20 + 20 + 20. No payment protocol (0). Hopper and IDA are paid products from other vendors and are optional, since Ghidra is free. Total 60.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-payments):

The published rubric, also on the [x402 page](https://www.anchorterminal.com/x402/).

- 40, a machine payment protocol (x402, MPP or L402) on the tool's own endpoints. 10 to 30 when it covers only some endpoints or only goes through a third party, and the note says which.
- 20, per-call or per-unit pricing published without a login. 10 for public plan-only pricing, 0 for "contact sales" or prices behind a login.
- 20, a free tier or trial that doesn't need a card.
- 20, autonomous onboarding, meaning an agent can get access without a person signing up in a browser (keyless use, x402, a programmatic key API).

Payment platforms and agent wallets rarely charge for their own API over a machine protocol, so the first line has steps for them, and the highest one that applies counts. 40 when x402, MPP or L402 runs on all their own endpoints, 30 when it runs on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, and 0 when the only protocol is their own. Merchant acceptance sits above a facilitator because the platform's own customers can charge agents through it, while a facilitator settles for sellers who wire up the protocol themselves. The counter-argument (a facilitator does more for the protocol as a whole) has a point. Each note says which step applied.

Open-source software you run yourself is scored on its hosted or paid option if it has one. A free, self-hosted package with nothing to buy gets 20, 20 and 20 for the last three lines, and 0 to 40 for the first only if it ships a payment protocol.

## 3. Reliability, 77 out of 100, up to 4.6 more on the total

Why it scored 77: Scored as a local stdio package. Official npm package rea-agents with Node.js 22.19+, 24.11+ or 26+ stated (20). The CI workflow runs on every push to main and every pull request, with sharded tests over 897 test files, a Windows native build and package checks. GitHub's Actions page served us an older copy, so we couldn't see whether main passes (20 of 25). The issue list we were served was also older than the repository header (55 open against 104), with three bug reports among 25 shown, all about Hopper. We couldn't read current issues, so 15 of 25. Breaking changes are listed under every release, but the always-bump-minor policy ships them in minor versions such as 6.2, 6.3 and 6.4 (7 of 15). Version 6.4.0 (15). Total 77.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-reliability):

Hosted APIs, MCP servers, models and platforms.

- 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own).
- 0 to 30, the incident record for the last 90 days on that page. 30 for a clean record or trivial incidents only, 20 for minor incidents only, 10 for one major outage (an hour or more of a core API down, or errors across the board), 0 for several. 5 when there's no history we could read, and the note says so.
- 15, rate limits documented with numbers.
- 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved.
- 10, an SLA published for any paid tier.
- 10, the surface agents use is generally available, not beta or preview.

Local packages, SDKs, frameworks and stdio MCP servers.

- 20, installs from an official package with supported runtimes stated.
- 25, a public CI and test suite, passing on the default branch.
- 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered).
- 15, semver discipline and breaking changes called out in a changelog.
- 15, version 1.0 or later, or declared stable.

Protocols are read from their reference implementations, the public facilitators or servers, spec stability and test vectors.

## 4. Agent ergonomics, 73 out of 100, up to 4.4 more on the total

Why it scored 73: 172 tools in the effect audit, and tools/list returns all of them even when a provider is missing (5). A compact schema profile shrinks each input schema below 13,312 bytes for size-capped clients, but there are no toolsets, dynamic loading or read-only subset, so no add-back. Large results return `resource_constraint` with a ready-to-call `inspect_analysis_view` continuation and `next_offset` paging, and `get_evidence_bundle` takes `detail: summary` with filters (18 of 20). Typed errors with category, retryable and remediation (20). Every tool's readOnlyHint, destructiveHint and idempotentHint come from an audited effect record (20). One command sets up 14 clients and the CLI mirrors the tools, but it is TypeScript only and many tools need long structured inputs (10 of 15). Total 73.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-ergonomics):

- 0 to 25, context cost. For MCP, the number and size of the tool definitions (25 for ten or fewer compact tools, 15 for 11 to 30, 5 for more than 30, plus up to 10 back for toolsets, dynamic loading or read-only subsets). For APIs, whether responses can be sized (field selection, limits, summaries).
- 20, pagination, filtering and output-size controls.
- 20, actionable, documented error responses, codes and messages an agent can recover from.
- 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations.
- 15, sensible defaults, few required parameters, and official SDKs in at least two languages.

Models are read for tool use, structured output, prompt caching, context length, batch and SDKs. Frameworks for how much code and how many defaults a tool-calling agent with MCP needs.

## 5. Transparency & trust, 53 out of 100, up to 4.1 more on the total

Made of editorial 78, provenance 27.

Why it scored 53: MIT (30). Local software. The README says targets are analysed locally and the model provider's own policy applies to tool results. SECURITY.md describes what the Ghidra session keeps and deletes, and browser captures drop cookies and authorisation headers. No privacy statement beyond that (20 of 30). Breaking changes are dated in the changelog with a 6.2 migration guide, but the policy says minor versions carry no compatibility promise and security fixes go to the latest only (8 of 20). No telemetry in the source, which names no outside host except github.com and Hopper's download origin. There's no written statement (20). Total 78.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-transparency):

- 0 to 30, source availability and licence clarity. 30 for open source under an OSI licence, 15 for closed with clear terms, 0 for unclear terms.
- 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors).
- 0 to 20, a deprecation policy or notices with dates.
- 0 to 20, telemetry disclosed with an opt-out (local software), or subprocessors and data locations disclosed (hosted).

The other half of Transparency and trust is the provenance score, computed from checked facts (below). The category score is the mean of the two.

Provenance checks not met in full (half of this category, computed from checked facts):

- Legal entity named: not found (0 of 20)
- Domain age: rea.tools, no registry record we could read (0 of 15)
- Status page: not found (0 of 10)
- security.txt: not found (0 of 10)

## 6. Schema & documentation, 86 out of 100, up to 2.3 more on the total

Why it scored 86: Every tool advertises typed JSON input and output schemas with no recursive references, per docs/mcp-contracts.md (25). No llms.txt on rea.tools. Markdown guides in docs/ and an agent skill installed by setup (5). Descriptions are long and specific. `open_binary` explains formats, provider binding and snapshot rules, and some say when to use them, such as `procedure_assembly` (15 of 20). Inputs are typed, but Evidence passed between tools travels as unconstrained JSON objects (11 of 15). An error-contract JSON Schema with codes, retryable and remediation fields, a user-visible error table and contract examples (15). Release Please versions with a dated changelog and a migration guide (15). Total 86.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-schema):

APIs and MCP servers.

- 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool).
- 10, llms.txt or Markdown docs served for agents.
- 0 to 20, descriptions that say what a tool is for, when to use it and when not to, read from the tool definitions in the source or the API reference.
- 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs.
- 0 to 15, examples and documented error responses.
- 15, versioning and a public changelog.

Models are read from the API reference, the OpenAPI file, llms.txt, the structured-output and tool-use docs and the model cards. Frameworks from docs a model can follow, typed interfaces, examples and the API reference.

## 7. Maintenance & community, 88 out of 100, up to 1.1 more on the total

Why it scored 88: 6.4.0 on 11 October 2026, one day before this check (30). 23 npm versions since 14 July 2026, nine of them between 3 and 11 October (20). The 6.4.0 notes record 336 commits and 163 merged pull requests from outside contributors, and the 6.x changelog closes 27 issues. We couldn't read current reply times on issues (15 of 25). In the official MCP registry as io.github.morluto/rea, latest 6.4.0 (15). Exact-pinned dependencies, MCP SDK 2.3.1, CI on push, no Dependabot config (8 of 10). Total 88.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-maintenance):

- 0 to 30, time since the last release, or the last published model or API change for a closed service. 30 within 30 days, 20 within 90, 10 within 180, 0 older.
- 20, at least three releases or dated changelog entries in the last 90 days.
- 0 to 25, responsiveness. Issues and pull requests answered on GitHub (the open issues and how recent the replies are). For closed services, a public changelog and a support or community channel that answers, 0 to 15.
- 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models).
- 10, package health, current dependencies and CI.

Models are read for deprecation notice periods and model churn rather than release counts.

## What we couldn't check

What we couldn't read counted as absent. Publishing it on a page a plain HTTP fetch can read (not only in a browser) lets the next check count it.

- unchecked: whether CI passes on main. GitHub's Actions page served a copy from around PR #1213, well before 6.4.0.
- unchecked: current open issues and reply times. The issues page we were served listed 55 open issues up to #497, while the repository header showed 104 and the changelog cites issues past #1400.
- The repository header showed 98.2k stars and 21.8k forks for a project first published on npm on 12 July 2026. The README celebrates 90,000. We recorded the header figure and did not assess how the stars accrued.
- No capability key in data/categories.json covers reverse engineering. The listing uses browser.debug and browser.control for its website tools, and its main job (binary and app analysis) has no key.
- No legal entity is named. The project is maintained by the GitHub account morluto.
- The server was not run, so the tool count (172) comes from the effect audit in source and the context cost of the tool list was not measured.

## Weaknesses

- 172 tools are advertised at once, including those whose provider is not installed, with no toolsets or read-only subset
- Breaking changes ship in minor versions, and 6.0 to 6.4 were released between 8 and 11 October 2026
- No per-call approval. Process, browser, Electron and Android install or uninstall tools run as the user once the client allows them
- Text recovered from binaries, bundles and web pages reaches the model with no documented prompt-injection guidance
- Deep native analysis depends on Hopper, Ghidra or IDA being installed, and Windows Ghidra support is marked experimental

## What costs an agent a turn today

The notes we give agents before they call it. Each one is a workaround an agent shouldn't need.

- Pin an exact version such as rea-agents@6.4.0 in the MCP config, because minor releases change result shapes
- Read `structuredContent.normalized_result` and `evidence_id` from results, and check `isError` before reading success fields
- When a result returns `resource_constraint`, call the `inspect_analysis_view` request in `details.reported_limits.continuation` and follow `next_offset`
- Call `binary_session` to see which tools and providers are available before opening a native target
- Set `REA_MCP_INPUT_SCHEMA_PROFILE=compact` for clients that reject large tool schemas, such as Kimi

## When it's done

Send what changed and where it's published as a dispute (https://www.anchorterminal.com/builders/#disputes, or `POST https://www.anchorterminal.com/api/v1/contact` with `"kind": "dispute"`). Disputes are answered in public, and the listing is checked again by the same checklist. Paying for an audit or a listing claim changes nothing here.