{
  "fixes": {
    "slug": "sourcegraph-mcp",
    "name": "Sourcegraph MCP Server",
    "listing": "https://www.anchorterminal.com/tools/sourcegraph-mcp",
    "markdown": "# Fix list: Sourcegraph MCP Server\n\nFrom Anchor Terminal's listing at https://www.anchorterminal.com/tools/sourcegraph-mcp, the October 2026 research run, assessed 8 October 2026. Grade C, 57.5 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 Sourcegraph MCP Server: 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, 25 out of 100, up to 15 more on the total\n\nWhy it scored 25: Scored with the hosted lines, because the server is an HTTP endpoint on a Sourcegraph Cloud or self-hosted instance. The site links a status page at sourcegraphstatus.com, but the host did not resolve from our network on 8 October 2026, so the page and its history are unread and scored as absent (0 + 0). This is our fetch failure, not an established gap. No request rate limits for the MCP endpoints were found in the reviewed documentation, only result limits per tool (0). No 429 or backoff guidance was found. Release 8.0.0 says search tools now separate incomplete searches from empty results and recommend a retry, and every tool reads, so retries are safe (5). The SLA page commits to 99.5 per cent monthly uptime for Sourcegraph Cloud on Enterprise plans (10). The server carries no beta label, and Code Finder became generally available on 20 August 2026 (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, 15 out of 100, up to 10.6 more on the total\n\nWhy it scored 15: No x402, MPP or L402 (0). The pricing page publishes a starting figure, a $16K minimum annual contract, with no per-user or per-call price and a contact-sales button, so half of the plan-pricing credit (5). A free Sourcegraph Cloud trial exists by request, with eligibility reviewed and a wait of up to an hour in business hours. No card is mentioned, so half credit (10). An agent cannot gain access alone. A company needs a contract or an approved trial, and a user approves OAuth in a browser (0).\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. Schema \u0026 documentation, 71 out of 100, up to 4.7 more on the total\n\nWhy it scored 71: The docs list each tool's parameters with required and optional marks, defaults and maximums. We could not read the tool definitions themselves, because every endpoint needs an Enterprise instance and sourcegraph.com answered 401, so the machine-readable contract earns partial credit (15). /llms.txt and /docs/llms.txt return 404, but every docs page is served as Markdown at its URL plus .md (10). Tool entries state use cases, `code_finder` has best practices and a contrast with `deepsearch`, and a matrix shows which endpoint carries which tool (14). Required fields, defaults and bounds are documented, while search filters travel inside one query string and `evaluator` takes a free Lua script (9). `list_refs` has JSON examples and the authentication page has a troubleshooting table for `invalid_scope`, 401, 403 and 404, with no catalogue of tool errors (8). Versioned releases with dated MCP entries in the changelog (15).\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## 4. Agent ergonomics, 72 out of 100, up to 4.6 more on the total\n\nWhy it scored 72: Sixteen documented tools on /.api/mcp/all, nine on the default endpoint and two on /.api/mcp/deepsearch, and admins can remove tools with `mcp.tools.disabled`. Definition sizes are unread (23). Result limits on every list and search tool, an `after` cursor on `compare_revisions`, `count` on searches and line ranges on `read_file` (17). Search tools report incomplete results with retry advice per the 8.0.0 notes, and HTTP 401, 403 and 404 causes are documented, but no tool error list was found (10). No tool writes to a repository, so repeating a call is safe. `deepsearch` creates a conversation each time, and we could not read the readOnlyHint or destructiveHint annotations (10). Most tools need one to three parameters with stated defaults. There is no SDK to count, as the surface is MCP only (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. Security \u0026 auth, 76 out of 100, up to 4.2 more on the total\n\nWhy it scored 76: OAuth 2.0 with PKCE, dynamic client registration held to the `mcp` scope, one-hour access tokens, a revocation endpoint and bearer tokens in the header only. Access tokens can also carry the `mcp` scope and default to a 90-day expiry (30). Every documented tool reads, repository permissions apply to each call, and admins can gate MCP by RBAC role or disable single tools. There is no per-repository grant narrower than the user's own access (18). The tools return repository content, which is untrusted input. The consent screen warns that the `mcp` scope grants read access to private code, and no prompt-injection guidance was found (4). The audit log records access token use, repository access and API requests as structured logs for a SIEM. No MCP-specific call log is documented (9). SOC 2 and ISO/IEC 27001:2022 on the security portal, annual third-party penetration tests, CVE fixes named in release notes and an invite-only HackerOne programme reached through security@sourcegraph.com. security.txt returns 404 (15).\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## 6. Transparency \u0026 trust, 78 out of 100, up to 1.9 more on the total\n\nMade of editorial 65, provenance 90.\n\nWhy it scored 78: Closed source under the Sourcegraph Terms of Service, last modified 22 July 2026, with supplemental terms listed in one index (15). The terms, the DPA, the security page and the privacy policy describe data handling. Cloud logs are kept up to 365 days, partner LLMs keep inputs only for abuse detection, and the privacy policy excludes customer code, which falls under the customer agreement. No retention period for code or search queries on Cloud was found (22). The current and previous major versions are supported, and release notes date removals such as the end of GitHub Enterprise Server 3.3 support, but no deprecation policy with a notice period for MCP tools was found (10). Sub-processors are listed with locations and purposes, last modified 21 August 2026, with email notice of changes, and instance telemetry is documented (18).\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- security.txt: not found (0 of 10)\n\n## 7. Maintenance \u0026 community, 83 out of 100, up to 1.5 more on the total\n\nWhy it scored 83: The latest Cloud update notes are dated 5 October 2026, and self-hosted 8.0.0 shipped on 17 September 2026 with a new `list_refs` tool (30). Four self-hosted releases since 10 July 2026 (7.6.0, 7.7.0, 7.7.359, 8.0.0) plus weekly update notes (20). A closed service with a public changelog, support at support@sourcegraph.com and published response times from two business hours for Enterprise. No public issue tracker for the server was found (12). Listed in the official MCP registry as io.github.sourcegraph/mcp (15). Release 7.6.0 moved to the MCP Go SDK v1.7.0 and protocol version 2026-07-28, and releases name the CVEs they fix. No public CI to read (6).\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: sourcegraphstatus.com did not resolve from our network (WebFetch and curl both failed), so the status page and its incident history are unread and scored as absent.\n- unchecked: the MCP tool definitions (JSON Schema inputs, description text, readOnlyHint and destructiveHint annotations, total size). Every endpoint needs an Enterprise instance, and sourcegraph.com/.api/mcp answered 401.\n- unchecked: the answers in the pricing page FAQ, among them the one on a free trial, are loaded on click and were not in the page we read. The trial facts come from the Cloud docs.\n- unchecked: the credit rate card for `code_finder` and Deep Search, which the changelog places in the Enterprise Portal.\n- The registry entry io.github.sourcegraph/mcp points at a repository under github.com/sourcegraph-community, which we did not open.\n- No request rate limit for the MCP endpoints was found. The security page says Cloudflare rate limiting protects Cloud domains, without numbers.\n- The lead was right on the endpoint, the authentication methods and the Enterprise-only restriction.\n\n## Weaknesses\n\n- Enterprise plans only, from a $16,000 minimum annual contract through sales. A Cloud trial is by request and subject to eligibility\n- No request rate limits, 429 behaviour or tool error catalogue found in the reviewed documentation\n- `code_finder` and Deep Search draw on the instance's credit entitlement and return an error once the quota is used up\n- No guidance on prompt injection from repository content was found, although every search and file tool returns such content\n- sourcegraph.com/.well-known/security.txt returns 404, and the bug bounty on HackerOne is invite-only\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- Connect to https://\u003cinstance\u003e/.api/mcp for the nine core tools. Use /.api/mcp/all for `go_to_definition`, `find_references`, `nls_search` and `compare_revisions`\n- Request only the `mcp` scope in OAuth. Any other scope on the MCP resource fails with `invalid_scope`\n- Without OAuth, send `Authorization: token \u003caccess token\u003e` and create the token with the `mcp` scope and an expiry\n- Call `list_repos` first and name the repository in every `code_finder` task. It declines broad searches across repositories\n- Read large files with `startLine` and `endLine`. Files over 128KB return only the first 200 lines\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": "C",
    "score": 57.5,
    "assessed": "2026-10-08",
    "run": "October 2026 research run",
    "categories": [
      {
        "key": "reliability",
        "name": "Reliability",
        "score": 25,
        "maxGain": 15,
        "reason": "Scored with the hosted lines, because the server is an HTTP endpoint on a Sourcegraph Cloud or self-hosted instance. The site links a status page at sourcegraphstatus.com, but the host did not resolve from our network on 8 October 2026, so the page and its history are unread and scored as absent (0 + 0). This is our fetch failure, not an established gap. No request rate limits for the MCP endpoints were found in the reviewed documentation, only result limits per tool (0). No 429 or backoff guidance was found. Release 8.0.0 says search tools now separate incomplete searches from empty results and recommend a retry, and every tool reads, so retries are safe (5). The SLA page commits to 99.5 per cent monthly uptime for Sourcegraph Cloud on Enterprise plans (10). The server carries no beta label, and Code Finder became generally available on 20 August 2026 (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": 15,
        "maxGain": 10.6,
        "reason": "No x402, MPP or L402 (0). The pricing page publishes a starting figure, a $16K minimum annual contract, with no per-user or per-call price and a contact-sales button, so half of the plan-pricing credit (5). A free Sourcegraph Cloud trial exists by request, with eligibility reviewed and a wait of up to an hour in business hours. No card is mentioned, so half credit (10). An agent cannot gain access alone. A company needs a contract or an approved trial, and a user approves OAuth in a browser (0).",
        "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": "schema",
        "name": "Schema \u0026 documentation",
        "score": 71,
        "maxGain": 4.7,
        "reason": "The docs list each tool's parameters with required and optional marks, defaults and maximums. We could not read the tool definitions themselves, because every endpoint needs an Enterprise instance and sourcegraph.com answered 401, so the machine-readable contract earns partial credit (15). /llms.txt and /docs/llms.txt return 404, but every docs page is served as Markdown at its URL plus .md (10). Tool entries state use cases, `code_finder` has best practices and a contrast with `deepsearch`, and a matrix shows which endpoint carries which tool (14). Required fields, defaults and bounds are documented, while search filters travel inside one query string and `evaluator` takes a free Lua script (9). `list_refs` has JSON examples and the authentication page has a troubleshooting table for `invalid_scope`, 401, 403 and 404, with no catalogue of tool errors (8). Versioned releases with dated MCP entries in the changelog (15).",
        "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": "ergonomics",
        "name": "Agent ergonomics",
        "score": 72,
        "maxGain": 4.6,
        "reason": "Sixteen documented tools on /.api/mcp/all, nine on the default endpoint and two on /.api/mcp/deepsearch, and admins can remove tools with `mcp.tools.disabled`. Definition sizes are unread (23). Result limits on every list and search tool, an `after` cursor on `compare_revisions`, `count` on searches and line ranges on `read_file` (17). Search tools report incomplete results with retry advice per the 8.0.0 notes, and HTTP 401, 403 and 404 causes are documented, but no tool error list was found (10). No tool writes to a repository, so repeating a call is safe. `deepsearch` creates a conversation each time, and we could not read the readOnlyHint or destructiveHint annotations (10). Most tools need one to three parameters with stated defaults. There is no SDK to count, as the surface is MCP only (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": "security",
        "name": "Security \u0026 auth",
        "score": 76,
        "maxGain": 4.2,
        "reason": "OAuth 2.0 with PKCE, dynamic client registration held to the `mcp` scope, one-hour access tokens, a revocation endpoint and bearer tokens in the header only. Access tokens can also carry the `mcp` scope and default to a 90-day expiry (30). Every documented tool reads, repository permissions apply to each call, and admins can gate MCP by RBAC role or disable single tools. There is no per-repository grant narrower than the user's own access (18). The tools return repository content, which is untrusted input. The consent screen warns that the `mcp` scope grants read access to private code, and no prompt-injection guidance was found (4). The audit log records access token use, repository access and API requests as structured logs for a SIEM. No MCP-specific call log is documented (9). SOC 2 and ISO/IEC 27001:2022 on the security portal, annual third-party penetration tests, CVE fixes named in release notes and an invite-only HackerOne programme reached through security@sourcegraph.com. security.txt returns 404 (15).",
        "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": "transparency",
        "name": "Transparency \u0026 trust",
        "score": 78,
        "maxGain": 1.9,
        "reason": "Closed source under the Sourcegraph Terms of Service, last modified 22 July 2026, with supplemental terms listed in one index (15). The terms, the DPA, the security page and the privacy policy describe data handling. Cloud logs are kept up to 365 days, partner LLMs keep inputs only for abuse detection, and the privacy policy excludes customer code, which falls under the customer agreement. No retention period for code or search queries on Cloud was found (22). The current and previous major versions are supported, and release notes date removals such as the end of GitHub Enterprise Server 3.3 support, but no deprecation policy with a notice period for MCP tools was found (10). Sub-processors are listed with locations and purposes, last modified 21 August 2026, with email notice of changes, and instance telemetry is documented (18).",
        "blend": "editorial 65, provenance 90",
        "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": "maintenance",
        "name": "Maintenance \u0026 community",
        "score": 83,
        "maxGain": 1.5,
        "reason": "The latest Cloud update notes are dated 5 October 2026, and self-hosted 8.0.0 shipped on 17 September 2026 with a new `list_refs` tool (30). Four self-hosted releases since 10 July 2026 (7.6.0, 7.7.0, 7.7.359, 8.0.0) plus weekly update notes (20). A closed service with a public changelog, support at support@sourcegraph.com and published response times from two business hours for Enterprise. No public issue tracker for the server was found (12). Listed in the official MCP registry as io.github.sourcegraph/mcp (15). Release 7.6.0 moved to the MCP Go SDK v1.7.0 and protocol version 2026-07-28, and releases name the CVEs they fix. No public CI to read (6).",
        "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": "security.txt",
        "value": "not found",
        "points": 0,
        "max": 10
      }
    ],
    "unchecked": [
      "unchecked: sourcegraphstatus.com did not resolve from our network (WebFetch and curl both failed), so the status page and its incident history are unread and scored as absent.",
      "unchecked: the MCP tool definitions (JSON Schema inputs, description text, readOnlyHint and destructiveHint annotations, total size). Every endpoint needs an Enterprise instance, and sourcegraph.com/.api/mcp answered 401.",
      "unchecked: the answers in the pricing page FAQ, among them the one on a free trial, are loaded on click and were not in the page we read. The trial facts come from the Cloud docs.",
      "unchecked: the credit rate card for `code_finder` and Deep Search, which the changelog places in the Enterprise Portal.",
      "The registry entry io.github.sourcegraph/mcp points at a repository under github.com/sourcegraph-community, which we did not open.",
      "No request rate limit for the MCP endpoints was found. The security page says Cloudflare rate limiting protects Cloud domains, without numbers.",
      "The lead was right on the endpoint, the authentication methods and the Enterprise-only restriction."
    ],
    "weaknesses": [
      "Enterprise plans only, from a $16,000 minimum annual contract through sales. A Cloud trial is by request and subject to eligibility",
      "No request rate limits, 429 behaviour or tool error catalogue found in the reviewed documentation",
      "`code_finder` and Deep Search draw on the instance's credit entitlement and return an error once the quota is used up",
      "No guidance on prompt injection from repository content was found, although every search and file tool returns such content",
      "sourcegraph.com/.well-known/security.txt returns 404, and the bug bounty on HackerOne is invite-only"
    ],
    "agentNotes": [
      "Connect to https://\u003cinstance\u003e/.api/mcp for the nine core tools. Use /.api/mcp/all for `go_to_definition`, `find_references`, `nls_search` and `compare_revisions`",
      "Request only the `mcp` scope in OAuth. Any other scope on the MCP resource fails with `invalid_scope`",
      "Without OAuth, send `Authorization: token \u003caccess token\u003e` and create the token with the `mcp` scope and an expiry",
      "Call `list_repos` first and name the repository in every `code_finder` task. It declines broad searches across repositories",
      "Read large files with `startLine` and `endLine`. Files over 128KB return only the first 200 lines"
    ],
    "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"
  }
}