{
  "fixes": {
    "slug": "grep-ai",
    "name": "Grep AI",
    "listing": "https://www.anchorterminal.com/tools/grep-ai",
    "markdown": "# Fix list: Grep AI\n\nFrom Anchor Terminal's listing at https://www.anchorterminal.com/tools/grep-ai, the October 2026 research run, assessed 8 October 2026. Grade B, 64.4 out of 100.\n\nThis 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.\n\nFor a coding agent working on Grep AI: 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.\n\n## 1. Reliability, 60 out of 100, up to 8 more on the total\n\nWhy it scored 60: Graded on the hosted API and MCP server. status.grep.ai runs on incident.io with a 90-day history, but its only component is Website, with none for the API or the MCP server (15). One incident in the last 90 days, \"Grep.ai currently unavailable\" on 4 September 2026, resolved after 24 minutes, so minor (20). No rate limit numbers found in the docs or the OpenAPI document (0). 429 `rate_limited` carries `Retry-After`, and run creation takes an `Idempotency-Key` with documented replay rules (15). The FAQ mentions SLA guarantees for enterprise customers, and no SLA is published (0). API v2 and the MCP server carry no beta label. The SDKs are at 0.2.1 and PyPI classifies the Python one as Alpha (10).\n\nThe checklist (https://www.anchorterminal.com/benchmark/#checklist-reliability):\n\nHosted APIs, MCP servers, models and platforms.\n\n- 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own).\n- 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.\n- 15, rate limits documented with numbers.\n- 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved.\n- 10, an SLA published for any paid tier.\n- 10, the surface agents use is generally available, not beta or preview.\n\nLocal packages, SDKs, frameworks and stdio MCP servers.\n\n- 20, installs from an official package with supported runtimes stated.\n- 25, a public CI and test suite, passing on the default branch.\n- 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered).\n- 15, semver discipline and breaking changes called out in a changelog.\n- 15, version 1.0 or later, or declared stable.\n\nProtocols are read from their reference implementations, the public facilitators or servers, spec stability and test vectors.\n\n## 2. Payments \u0026 pricing, 40 out of 100, up to 7.5 more on the total\n\nWhy it scored 40: No machine payment protocol on the live endpoints. The changelog of 8 May 2026 announces x402 and MPP, and the gateway answered 503 `gateway_disabled` on 8 October 2026 (0). Plan and credit prices are public, but only the medium tier's credit cost is published and the page and its Markdown twin disagree on credits per dollar (10). 100 free trial credits, and 500 for an owner's first agent through AgentID. The pricing page doesn't say whether a card is needed, and its table marks the trial without API access (15). AgentID sign-in is documented as needing no human account, but it depends on an AgentMail inbox and a browser step, and we didn't run it (15).\n\nThe checklist (https://www.anchorterminal.com/benchmark/#checklist-payments):\n\nThe published rubric, also on the [x402 page](https://www.anchorterminal.com/x402/).\n\n- 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.\n- 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.\n- 20, a free tier or trial that doesn't need a card.\n- 20, autonomous onboarding, meaning an agent can get access without a person signing up in a browser (keyless use, x402, a programmatic key API).\n\nPayment 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.\n\nOpen-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.\n\n## 3. Security \u0026 auth, 65 out of 100, up to 6.1 more on the total\n\nWhy it scored 65: OAuth through Descope with five scopes, and API keys that can be created with a subset of scopes, listed and revoked through the API (30). Read scopes, `validate_only` and `dry_run` on agent changes, a per-run `max_cost_usd` and versioned restores. No confirmation step for deletes or for making a run public (14). Reports are built from untrusted web pages. The docs describe an SSRF-guarded fetch for context files, and no prompt-injection guidance for callers was found (3). Each run has a timeline of execution events, and billing transactions and usage are readable through the API (12). The footer and FAQ claim SOC 2 Type II with no report or trust page linked. security.txt returns 404, and no disclosure policy or bug bounty was found (6). Webhooks are signed with the API key itself.\n\nThe checklist (https://www.anchorterminal.com/benchmark/#checklist-security):\n\n- 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.\n- 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions.\n- 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.\n- 0 to 15, audit logs or per-call visibility for the operator.\n- 0 to 20, a security programme. security.txt or a disclosure policy, a bug bounty, SOC 2 or ISO 27001, advisories handled in public.\n\nModels 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.\n\n## 4. Agent ergonomics, 70 out of 100, up to 4.9 more on the total\n\nWhy it scored 70: The MCP server card lists 82 tools, 31 of them deprecated aliases, with no documented toolsets. Reports default to plain Markdown and large files come back as signed URLs, so some credit back (8). Cursor pagination with `limit` on runs, timelines and transactions, and server-side filters on the skill and tool lists (18). One error envelope with stable codes and a `request_id`, and 402 for credits kept apart from 429 for rate (20). `Idempotency-Key` with body-hash matching and a 24-hour replay window. The server card shows no readOnlyHint or destructiveHint annotations, and the live tool list wasn't read (12). Only `question` is required and `effort` defaults to low. TypeScript and Python SDKs exist, though one docs page still calls the release pending and names a PyPI package that returns 404 (12).\n\nThe checklist (https://www.anchorterminal.com/benchmark/#checklist-ergonomics):\n\n- 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).\n- 20, pagination, filtering and output-size controls.\n- 20, actionable, documented error responses, codes and messages an agent can recover from.\n- 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations.\n- 15, sensible defaults, few required parameters, and official SDKs in at least two languages.\n\nModels 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.\n\n## 5. Transparency \u0026 trust, 59 out of 100, up to 3.6 more on the total\n\nMade of editorial 36, provenance 82.\n\nWhy it scored 59: Closed service with published terms, written as an order-form contract that still points to grep.parcha.ai. The SDK packages state no licence, and the skills package is MIT (12). Section 12 of the terms says AI providers retain nothing and customer data isn't used for training. Section 4.2 grants a perpetual licence to use customer data to improve the platform, and the privacy policy gives retention only as \"as long as needed\". No DPA found (12). v1 is marked deprecated, /research answers with a `deprecation: true` header and a successor link, and 31 MCP tools are labelled deprecated, all without dates or a written policy (8). The privacy policy names Google Analytics and PostHog and says data may be processed in any country. No subprocessor list found (4).\n\nThe checklist (https://www.anchorterminal.com/benchmark/#checklist-transparency):\n\n- 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.\n- 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors).\n- 0 to 20, a deprecation policy or notices with dates.\n- 0 to 20, telemetry disclosed with an opt-out (local software), or subprocessors and data locations disclosed (hosted).\n\nThe other half of Transparency and trust is the provenance score, computed from checked facts (below). The category score is the mean of the two.\n\nProvenance checks not met in full (half of this category, computed from checked facts):\n\n- Domain age: grep.ai, registered 2018-04-20 (8 years) (11 of 15)\n- Terms of service: read, states 5 of the 7 things a reader expects, and has 1 clause that costs points (6.3 of 10)\n- security.txt: not found (0 of 10)\n\n## 6. Schema \u0026 documentation, 80 out of 100, up to 3.3 more on the total\n\nWhy it scored 80: Public OpenAPI document at api.grep.ai/api/v2/openapi.json with 126 operations and 145 schemas (25). llms.txt and a Markdown twin of each developer page (10). Most operations carry a description, and the `effort` field says what each tier returns and how long it takes. 19 operations for campaigns, loops and batch schedules have none, and the MCP server card gives each of its 82 tools one line, one of them a placeholder. The MCP input schemas need OAuth and weren't read (12). 58 enums, numeric bounds and `additionalProperties: false` on the run request (12). Curl examples and an 11-code error registry in the docs, but the OpenAPI document has no examples and documents only 200 and 422 responses (9). The path is versioned, v1 is marked deprecated and the changelog is dated, with no entry after 24 July 2026 (12).\n\nThe checklist (https://www.anchorterminal.com/benchmark/#checklist-schema):\n\nAPIs and MCP servers.\n\n- 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool).\n- 10, llms.txt or Markdown docs served for agents.\n- 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.\n- 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs.\n- 0 to 15, examples and documented error responses.\n- 15, versioning and a public changelog.\n\nModels 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.\n\n## 7. Maintenance \u0026 community, 74 out of 100, up to 2.3 more on the total\n\nWhy it scored 74: @parcha/agentrun 0.2.1 on npm and parcha-agentrun 0.2.1 on PyPI, both on 30 September 2026 (30). Three npm releases of the SDK since 21 September, grep-research-skills 0.2.0 on 4 August, and changelog entries on 11 and 24 July (20). Closed service with a public changelog whose newest entry is 24 July 2026, and support at support@parcha.ai. We didn't test whether support answers (8). Current SDKs in two languages, with docs that lag them. Not found in the official MCP registry (12). The skills repository runs a test workflow. The SDK repositories named in the package metadata asked for credentials, so their CI wasn't read (4).\n\nThe checklist (https://www.anchorterminal.com/benchmark/#checklist-maintenance):\n\n- 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.\n- 20, at least three releases or dated changelog entries in the last 90 days.\n- 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.\n- 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models).\n- 10, package health, current dependencies and CI.\n\nModels are read for deprecation notice periods and model churn rather than release counts.\n\n## What we couldn't check\n\nWhat 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.\n\n- unchecked: the MCP tool input schemas and annotations, because tools/list needs an OAuth sign-in. The server card gives names and one-line descriptions only\n- unchecked: whether Free Trial, Pay As You Go or AgentID accounts can call the API or MCP server. The pricing table says API access is Pro and Ultra only, and no account was opened\n- unchecked: whether signup needs a card\n- unchecked: GitHub stars and issue activity, because the GitHub API refused us for its rate limit. The SDK repositories named in the package metadata asked for credentials\n- unchecked: the SOC 2 Type II report, which the footer and FAQ claim without a link\n- Credits per dollar on Pay As You Go. The pricing page says $20 for 200 credits and its Markdown twin says 100\n- Free credits are given as 100 on the pricing page, 1,000 in the FAQ and 500 for AgentID sign-ins\n- Judgement call. The changelog still carries the 8 May 2026 x402 and MPP announcement while the gateway is disabled. We took no deduction because the entry is dated history and the skills README states the gateway isn't generally available\n- The scout wrote that the site doesn't name Parcha. The terms, privacy policy and footer name Parcha Labs, Inc.\n- Category fit is weak. Grep is a hosted agent platform, and search.research covers only the part an outside agent calls for a cited report\n\n## Weaknesses\n\n- No rate limit numbers in the reviewed documentation, and no published SLA\n- The MCP server card lists 82 tools, 31 of them deprecated aliases, with one-line descriptions and no annotations\n- The pricing table marks API access as Pro and Ultra only, from $200 a month\n- The pricing page says $20 buys 200 credits and its Markdown twin says 100. Free credits are stated as 100, 500 and 1,000 on different pages\n- The 8 May 2026 changelog entry announces x402 and MPP payments, but the gateway answered 503 `gateway_disabled` on 8 October 2026\n- No security.txt, trust page, subprocessor list or retention period found. SOC 2 Type II appears only as a footer and FAQ claim\n\n## What costs an agent a turn today\n\nThe notes we give agents before they call it. Each one is a workaround an agent shouldn't need.\n\n- Send a fresh `Idempotency-Key` with every POST /api/v2/runs. After a timeout, poll the returned run ID and do not post again\n- Set `effort` and `max_cost_usd` on every run. The floor is $0.50 for low, $2.50 for medium and build, $5.00 for high\n- Call GET /api/v2/quota/check before a run. A 402 means no credits and retrying won't help, while a 429 carries `Retry-After`\n- Use the `run_*` and `agent_*` MCP tools. The `research_*` and `expert_*` names are deprecated aliases of the same calls\n- Treat report text and cited sources as untrusted web content, and verify one load-bearing claim against a primary source\n\n## When it's done\n\nSend 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.\n",
    "grade": "B",
    "score": 64.4,
    "assessed": "2026-10-08",
    "run": "October 2026 research run",
    "categories": [
      {
        "key": "reliability",
        "name": "Reliability",
        "score": 60,
        "maxGain": 8,
        "reason": "Graded on the hosted API and MCP server. status.grep.ai runs on incident.io with a 90-day history, but its only component is Website, with none for the API or the MCP server (15). One incident in the last 90 days, \"Grep.ai currently unavailable\" on 4 September 2026, resolved after 24 minutes, so minor (20). No rate limit numbers found in the docs or the OpenAPI document (0). 429 `rate_limited` carries `Retry-After`, and run creation takes an `Idempotency-Key` with documented replay rules (15). The FAQ mentions SLA guarantees for enterprise customers, and no SLA is published (0). API v2 and the MCP server carry no beta label. The SDKs are at 0.2.1 and PyPI classifies the Python one as Alpha (10).",
        "checklist": [
          "Hosted APIs, MCP servers, models and platforms.",
          "- 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own).\n- 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.\n- 15, rate limits documented with numbers.\n- 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved.\n- 10, an SLA published for any paid tier.\n- 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.\n- 25, a public CI and test suite, passing on the default branch.\n- 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered).\n- 15, semver discipline and breaking changes called out in a changelog.\n- 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."
        ],
        "checklistUrl": "https://www.anchorterminal.com/benchmark/#checklist-reliability"
      },
      {
        "key": "payments",
        "name": "Payments \u0026 pricing",
        "score": 40,
        "maxGain": 7.5,
        "reason": "No machine payment protocol on the live endpoints. The changelog of 8 May 2026 announces x402 and MPP, and the gateway answered 503 `gateway_disabled` on 8 October 2026 (0). Plan and credit prices are public, but only the medium tier's credit cost is published and the page and its Markdown twin disagree on credits per dollar (10). 100 free trial credits, and 500 for an owner's first agent through AgentID. The pricing page doesn't say whether a card is needed, and its table marks the trial without API access (15). AgentID sign-in is documented as needing no human account, but it depends on an AgentMail inbox and a browser step, and we didn't run it (15).",
        "checklist": [
          "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.\n- 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.\n- 20, a free tier or trial that doesn't need a card.\n- 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."
        ],
        "checklistUrl": "https://www.anchorterminal.com/benchmark/#checklist-payments"
      },
      {
        "key": "security",
        "name": "Security \u0026 auth",
        "score": 65,
        "maxGain": 6.1,
        "reason": "OAuth through Descope with five scopes, and API keys that can be created with a subset of scopes, listed and revoked through the API (30). Read scopes, `validate_only` and `dry_run` on agent changes, a per-run `max_cost_usd` and versioned restores. No confirmation step for deletes or for making a run public (14). Reports are built from untrusted web pages. The docs describe an SSRF-guarded fetch for context files, and no prompt-injection guidance for callers was found (3). Each run has a timeline of execution events, and billing transactions and usage are readable through the API (12). The footer and FAQ claim SOC 2 Type II with no report or trust page linked. security.txt returns 404, and no disclosure policy or bug bounty was found (6). Webhooks are signed with the API key itself.",
        "checklist": [
          "- 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.\n- 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions.\n- 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.\n- 0 to 15, audit logs or per-call visibility for the operator.\n- 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."
        ],
        "checklistUrl": "https://www.anchorterminal.com/benchmark/#checklist-security"
      },
      {
        "key": "ergonomics",
        "name": "Agent ergonomics",
        "score": 70,
        "maxGain": 4.9,
        "reason": "The MCP server card lists 82 tools, 31 of them deprecated aliases, with no documented toolsets. Reports default to plain Markdown and large files come back as signed URLs, so some credit back (8). Cursor pagination with `limit` on runs, timelines and transactions, and server-side filters on the skill and tool lists (18). One error envelope with stable codes and a `request_id`, and 402 for credits kept apart from 429 for rate (20). `Idempotency-Key` with body-hash matching and a 24-hour replay window. The server card shows no readOnlyHint or destructiveHint annotations, and the live tool list wasn't read (12). Only `question` is required and `effort` defaults to low. TypeScript and Python SDKs exist, though one docs page still calls the release pending and names a PyPI package that returns 404 (12).",
        "checklist": [
          "- 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).\n- 20, pagination, filtering and output-size controls.\n- 20, actionable, documented error responses, codes and messages an agent can recover from.\n- 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations.\n- 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."
        ],
        "checklistUrl": "https://www.anchorterminal.com/benchmark/#checklist-ergonomics"
      },
      {
        "key": "transparency",
        "name": "Transparency \u0026 trust",
        "score": 59,
        "maxGain": 3.6,
        "reason": "Closed service with published terms, written as an order-form contract that still points to grep.parcha.ai. The SDK packages state no licence, and the skills package is MIT (12). Section 12 of the terms says AI providers retain nothing and customer data isn't used for training. Section 4.2 grants a perpetual licence to use customer data to improve the platform, and the privacy policy gives retention only as \"as long as needed\". No DPA found (12). v1 is marked deprecated, /research answers with a `deprecation: true` header and a successor link, and 31 MCP tools are labelled deprecated, all without dates or a written policy (8). The privacy policy names Google Analytics and PostHog and says data may be processed in any country. No subprocessor list found (4).",
        "blend": "editorial 36, provenance 82",
        "checklist": [
          "- 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.\n- 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors).\n- 0 to 20, a deprecation policy or notices with dates.\n- 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."
        ],
        "checklistUrl": "https://www.anchorterminal.com/benchmark/#checklist-transparency"
      },
      {
        "key": "schema",
        "name": "Schema \u0026 documentation",
        "score": 80,
        "maxGain": 3.3,
        "reason": "Public OpenAPI document at api.grep.ai/api/v2/openapi.json with 126 operations and 145 schemas (25). llms.txt and a Markdown twin of each developer page (10). Most operations carry a description, and the `effort` field says what each tier returns and how long it takes. 19 operations for campaigns, loops and batch schedules have none, and the MCP server card gives each of its 82 tools one line, one of them a placeholder. The MCP input schemas need OAuth and weren't read (12). 58 enums, numeric bounds and `additionalProperties: false` on the run request (12). Curl examples and an 11-code error registry in the docs, but the OpenAPI document has no examples and documents only 200 and 422 responses (9). The path is versioned, v1 is marked deprecated and the changelog is dated, with no entry after 24 July 2026 (12).",
        "checklist": [
          "APIs and MCP servers.",
          "- 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool).\n- 10, llms.txt or Markdown docs served for agents.\n- 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.\n- 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs.\n- 0 to 15, examples and documented error responses.\n- 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."
        ],
        "checklistUrl": "https://www.anchorterminal.com/benchmark/#checklist-schema"
      },
      {
        "key": "maintenance",
        "name": "Maintenance \u0026 community",
        "score": 74,
        "maxGain": 2.3,
        "reason": "@parcha/agentrun 0.2.1 on npm and parcha-agentrun 0.2.1 on PyPI, both on 30 September 2026 (30). Three npm releases of the SDK since 21 September, grep-research-skills 0.2.0 on 4 August, and changelog entries on 11 and 24 July (20). Closed service with a public changelog whose newest entry is 24 July 2026, and support at support@parcha.ai. We didn't test whether support answers (8). Current SDKs in two languages, with docs that lag them. Not found in the official MCP registry (12). The skills repository runs a test workflow. The SDK repositories named in the package metadata asked for credentials, so their CI wasn't read (4).",
        "checklist": [
          "- 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.\n- 20, at least three releases or dated changelog entries in the last 90 days.\n- 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.\n- 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models).\n- 10, package health, current dependencies and CI.",
          "Models are read for deprecation notice periods and model churn rather than release counts."
        ],
        "checklistUrl": "https://www.anchorterminal.com/benchmark/#checklist-maintenance"
      }
    ],
    "provenance": [
      {
        "label": "Domain age",
        "value": "grep.ai, registered 2018-04-20 (8 years)",
        "points": 11,
        "max": 15
      },
      {
        "label": "Terms of service",
        "value": "read, states 5 of the 7 things a reader expects, and has 1 clause that costs points",
        "points": 6.3,
        "max": 10
      },
      {
        "label": "security.txt",
        "value": "not found",
        "points": 0,
        "max": 10
      }
    ],
    "unchecked": [
      "unchecked: the MCP tool input schemas and annotations, because tools/list needs an OAuth sign-in. The server card gives names and one-line descriptions only",
      "unchecked: whether Free Trial, Pay As You Go or AgentID accounts can call the API or MCP server. The pricing table says API access is Pro and Ultra only, and no account was opened",
      "unchecked: whether signup needs a card",
      "unchecked: GitHub stars and issue activity, because the GitHub API refused us for its rate limit. The SDK repositories named in the package metadata asked for credentials",
      "unchecked: the SOC 2 Type II report, which the footer and FAQ claim without a link",
      "Credits per dollar on Pay As You Go. The pricing page says $20 for 200 credits and its Markdown twin says 100",
      "Free credits are given as 100 on the pricing page, 1,000 in the FAQ and 500 for AgentID sign-ins",
      "Judgement call. The changelog still carries the 8 May 2026 x402 and MPP announcement while the gateway is disabled. We took no deduction because the entry is dated history and the skills README states the gateway isn't generally available",
      "The scout wrote that the site doesn't name Parcha. The terms, privacy policy and footer name Parcha Labs, Inc.",
      "Category fit is weak. Grep is a hosted agent platform, and search.research covers only the part an outside agent calls for a cited report"
    ],
    "weaknesses": [
      "No rate limit numbers in the reviewed documentation, and no published SLA",
      "The MCP server card lists 82 tools, 31 of them deprecated aliases, with one-line descriptions and no annotations",
      "The pricing table marks API access as Pro and Ultra only, from $200 a month",
      "The pricing page says $20 buys 200 credits and its Markdown twin says 100. Free credits are stated as 100, 500 and 1,000 on different pages",
      "The 8 May 2026 changelog entry announces x402 and MPP payments, but the gateway answered 503 `gateway_disabled` on 8 October 2026",
      "No security.txt, trust page, subprocessor list or retention period found. SOC 2 Type II appears only as a footer and FAQ claim"
    ],
    "agentNotes": [
      "Send a fresh `Idempotency-Key` with every POST /api/v2/runs. After a timeout, poll the returned run ID and do not post again",
      "Set `effort` and `max_cost_usd` on every run. The floor is $0.50 for low, $2.50 for medium and build, $5.00 for high",
      "Call GET /api/v2/quota/check before a run. A 402 means no credits and retrying won't help, while a 429 carries `Retry-After`",
      "Use the `run_*` and `agent_*` MCP tools. The `research_*` and `expert_*` names are deprecated aliases of the same calls",
      "Treat report text and cited sources as untrusted web content, and verify one load-bearing claim against a primary source"
    ],
    "recheck": "https://www.anchorterminal.com/builders/#disputes"
  },
  "meta": {
    "attribution": "Anchor Terminal (https://www.anchorterminal.com)",
    "docs": "https://www.anchorterminal.com/docs/",
    "generatedAt": "2026-10-08",
    "license": "CC-BY-4.0",
    "method": "https://www.anchorterminal.com/benchmark/",
    "methodology": "0.4",
    "openapi": "https://www.anchorterminal.com/openapi.json",
    "preview": false,
    "run": "2026-10-01",
    "runLabel": "October 2026 research run"
  }
}