{
  "fixes": {
    "slug": "prisma-airs",
    "name": "Prisma AIRS AI Runtime Security API",
    "listing": "https://www.anchorterminal.com/tools/prisma-airs",
    "markdown": "# Fix list: Prisma AIRS AI Runtime Security API\n\nFrom Anchor Terminal's listing at https://www.anchorterminal.com/tools/prisma-airs, the October 2026 research run, assessed 8 October 2026. Grade B, 62.8 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 Prisma AIRS AI Runtime Security API: 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. Payments \u0026 pricing, 0 out of 100, up to 12.5 more on the total\n\nWhy it scored 0: No x402, MPP or L402 (0). No money price is published. The credit estimator returns Software NGFW credits and sends buyers to sales for a quote (0). No free tier or self-serve trial found. The EULA allows 30-day evaluations by arrangement (0). Access needs a credit pool, a deployment profile in the Customer Support Portal and onboarding in Strata Cloud Manager by a person, and first activation of a tenant needs vendor support (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## 2. Maintenance \u0026 community, 31 out of 100, up to 6 more on the total\n\nWhy it scored 31: The last dated release we could source is `pan-aisecurity` 0.11.0 on PyPI on 15 May 2026, 146 days ago (10 of 30). The SDK repository's last commit is 20 July 2026 and the OpenAPI documents last changed on 5 March 2026. No dated releases or changelog entries in the last 90 days. Notes on new capabilities are on a Platform Explorer page drawn by script, which we couldn't read (0). No public changelog for the API, with support through the Customer Support Portal (5 of 15). Python SDKs for the scan and management APIs are current, in one language (10 of 15). The SDK has tests, a lock file, pre-commit checks and a CI task file, with CI results not public (6 of 10).\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## 3. Agent ergonomics, 65 out of 100, up to 5.7 more on the total\n\nWhy it scored 65: A scan answer is a set of boolean flags per detector plus `action` and `category`, with detail objects only where something matched, and no field selection (18 of 25). The security profile picks the detectors, management lists page with `offset` and `limit`, and results and reports take at most 5 IDs (12 of 20). Nine status codes with listed causes, but one message string and no machine-readable code, and detector failures arrive in a 200 with `error` or `timeout` set (13 of 20). Scans change nothing but the log and the quota, `tr_id` and `session_id` correlate calls, and the 429 body says when to retry. No idempotency key, and MCP tool annotations weren't read (14 of 20). Two required fields. Python is the only official SDK (8 of 15).\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## 4. Schema \u0026 documentation, 68 out of 100, up to 5.2 more on the total\n\nWhy it scored 68: OpenAPI 3.0.3 documents for the scan API (4 operations) and the management API (21) are public in the pan.dev repository (25). pan.dev/llms.txt returns 404, and pages have a copy-as-Markdown button with Markdown sources on GitHub (3 of 10). The overview and a use-case page explain each detector and that only the last element of `contents` is scanned, without saying when not to call (13 of 20). `contents` and `ai_profile` are required and detector names are an enum, but `category` and `action` are plain strings, tool input and output are raw JSON strings, and either-or rules for the profile sit in prose (9 of 15). Ten worked use cases with requests and responses and an error table with common causes per status code (13 of 15). The path is versioned `/v1`, but no dated API changelog was found, and the SDK's changelog stops at 1 December 2025 while PyPI has 0.11.0 (5 of 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## 5. Reliability, 87 out of 100, up to 2.6 more on the total\n\nWhy it scored 87: Read with the hosted lines, for the scan API. status.paloaltonetworks.com has a Prisma AIRS group with an AI Runtime Security API component (20). The incident feed from 10 July to 8 October 2026 has 23 entries, none naming a Prisma AIRS component (30). Results and reports are limited to 10 requests a minute and 5 IDs a call, and payloads to 2 MB and 5 MB, while scan volume is capped only by the monthly token quota, with no per-second figure (10 of 15). The 429 body carries `retry_after` with an interval and unit, and the Python SDK retries 500, 502, 503 and 504 five times by default, with no written backoff guidance (12 of 15). A Service Level Objective of 99.9 per cent per region on commercially reasonable efforts, dated 2 June 2026, and no SLA with remedies found (5 of 10). The docs carry no beta or preview label, though the specification's version field reads 0.0.0 (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## 6. Security \u0026 auth, 85 out of 100, up to 2.6 more on the total\n\nWhy it scored 85: API keys with a rotation period, expiry, regeneration and revocation, or OAuth 2.0 tokens from a service account. Keys are bound to one region and deployment profile and sent in a header (27 of 30). Strata Cloud Manager has a View Only Administrator role and custom roles with create, read, update and delete permissions per entity, and a scan key can't change configuration. Force-delete operations have no confirmation step (15 of 20). The service is itself an injection detector that scans tool inputs and outputs. Reports return the original payloads (13 of 15). An API scan log and log viewer in Strata Cloud Manager list each scan with its verdict, profile and action (15). PSIRT with a disclosure policy, a report form, a bug bounty and public advisories, and SOC 2+ on the certifications page without a product list we could read. No security.txt, and the onboarding sample curl passes `-k` (15 of 20).\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## 7. Transparency \u0026 trust, 71 out of 100, up to 2.5 more on the total\n\nMade of editorial 61, provenance 80.\n\nWhy it scored 71: Closed service under the EULA of August 2026, which names the contracting entity by region. The SDK is under the PolyForm Internal Use licence while the OpenAPI document's licence field says MIT (15 of 30). The privacy datasheet of 13 July 2026 lists the data processed, who can access it and retention (benign payloads up to 14 days, malicious up to 10 years), the EULA bars training generative AI on customer data, and a DPA is referenced. The 10-year period is disclosed, not hidden (26 of 30). No deprecation policy or dated notices for the API found, and the docs still mention a daily call quota alongside the monthly token model (0). GCP named as sub-processor with five countries, and out-of-region processing listed per detector (20).\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- Changelog: not found (0 of 10)\n- security.txt: not found (0 of 10)\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 Platform Explorer page that lists new capabilities (https://docs.paloaltonetworks.com/platform-explorer), which is drawn by script, so dated notes for the API in the last 90 days may exist.\n- unchecked: the MCP server's tool list and annotations, which need a key. The docs name `pan_inline_scan` and log a batch tool as `pan_batch_tool`.\n- unchecked: commit dates through the GitHub API, which was rate-limited for our address. Dates come from a shallow clone of PaloAltoNetworks/pan.dev.\n- Which certifications cover Prisma AIRS. The certifications page lists SOC 2+, FedRAMP, C5, IRAP and PCI DSS without a product list we could read, and the docs say the API intercept isn't available in FedRAMP environments.\n- The money price of a billion tokens a month, and whether failed scans count against the quota.\n- Whether the onboarding page's daily API call quota still applies to any customer, since the deployment profile page describes monthly tokens.\n- The EULA forbids publishing benchmark or comparison tests and load testing, which matters before any probe of ours runs.\n- `provenance.privacy` points at the product privacy datasheet, a PDF. The general Privacy Statement covers only data the company controls.\n\n## Weaknesses\n\n- No public price, free tier or trial. Capacity is bought as Software NGFW credits, in steps of 1 billion tokens a month, through sales\n- Payloads judged malicious are kept for up to 10 years, including after the subscription ends, per the privacy datasheet of 13 July 2026\n- The EULA of August 2026 forbids publishing benchmark or comparison tests and forbids load testing of subscriptions\n- Python is the only official SDK, and no dated changelog for the API was found\n- Uptime is a 99.9 per cent objective on commercially reasonable efforts, with no SLA or credits found for Prisma AIRS\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 `x-pan-token` and an `ai_profile` with `profile_name` or `profile_id` on every scan. Both come from Strata Cloud Manager, and a key works only in its own region\n- Put earlier turns first in `contents`. Only the last element is scanned, and the rest is context\n- Read `action` (allow or block) and `category`, then check `timeout` and `error`, because a detector that failed is reported in `errors` with a 200\n- Keep synchronous requests under 2 MB and asynchronous ones under 5 MB and 25 items, and fetch at most 5 scan or report IDs a call, 10 calls a minute\n- For MCP, connect to `/mcp` on the regional host with `x-pan-token` and `x-pan-profile` headers, and call `pan_inline_scan` yourself before and after generation\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": 62.8,
    "assessed": "2026-10-08",
    "run": "October 2026 research run",
    "categories": [
      {
        "key": "payments",
        "name": "Payments \u0026 pricing",
        "score": 0,
        "maxGain": 12.5,
        "reason": "No x402, MPP or L402 (0). No money price is published. The credit estimator returns Software NGFW credits and sends buyers to sales for a quote (0). No free tier or self-serve trial found. The EULA allows 30-day evaluations by arrangement (0). Access needs a credit pool, a deployment profile in the Customer Support Portal and onboarding in Strata Cloud Manager by a person, and first activation of a tenant needs vendor support (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": "maintenance",
        "name": "Maintenance \u0026 community",
        "score": 31,
        "maxGain": 6,
        "reason": "The last dated release we could source is `pan-aisecurity` 0.11.0 on PyPI on 15 May 2026, 146 days ago (10 of 30). The SDK repository's last commit is 20 July 2026 and the OpenAPI documents last changed on 5 March 2026. No dated releases or changelog entries in the last 90 days. Notes on new capabilities are on a Platform Explorer page drawn by script, which we couldn't read (0). No public changelog for the API, with support through the Customer Support Portal (5 of 15). Python SDKs for the scan and management APIs are current, in one language (10 of 15). The SDK has tests, a lock file, pre-commit checks and a CI task file, with CI results not public (6 of 10).",
        "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"
      },
      {
        "key": "ergonomics",
        "name": "Agent ergonomics",
        "score": 65,
        "maxGain": 5.7,
        "reason": "A scan answer is a set of boolean flags per detector plus `action` and `category`, with detail objects only where something matched, and no field selection (18 of 25). The security profile picks the detectors, management lists page with `offset` and `limit`, and results and reports take at most 5 IDs (12 of 20). Nine status codes with listed causes, but one message string and no machine-readable code, and detector failures arrive in a 200 with `error` or `timeout` set (13 of 20). Scans change nothing but the log and the quota, `tr_id` and `session_id` correlate calls, and the 429 body says when to retry. No idempotency key, and MCP tool annotations weren't read (14 of 20). Two required fields. Python is the only official SDK (8 of 15).",
        "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": "schema",
        "name": "Schema \u0026 documentation",
        "score": 68,
        "maxGain": 5.2,
        "reason": "OpenAPI 3.0.3 documents for the scan API (4 operations) and the management API (21) are public in the pan.dev repository (25). pan.dev/llms.txt returns 404, and pages have a copy-as-Markdown button with Markdown sources on GitHub (3 of 10). The overview and a use-case page explain each detector and that only the last element of `contents` is scanned, without saying when not to call (13 of 20). `contents` and `ai_profile` are required and detector names are an enum, but `category` and `action` are plain strings, tool input and output are raw JSON strings, and either-or rules for the profile sit in prose (9 of 15). Ten worked use cases with requests and responses and an error table with common causes per status code (13 of 15). The path is versioned `/v1`, but no dated API changelog was found, and the SDK's changelog stops at 1 December 2025 while PyPI has 0.11.0 (5 of 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": "reliability",
        "name": "Reliability",
        "score": 87,
        "maxGain": 2.6,
        "reason": "Read with the hosted lines, for the scan API. status.paloaltonetworks.com has a Prisma AIRS group with an AI Runtime Security API component (20). The incident feed from 10 July to 8 October 2026 has 23 entries, none naming a Prisma AIRS component (30). Results and reports are limited to 10 requests a minute and 5 IDs a call, and payloads to 2 MB and 5 MB, while scan volume is capped only by the monthly token quota, with no per-second figure (10 of 15). The 429 body carries `retry_after` with an interval and unit, and the Python SDK retries 500, 502, 503 and 504 five times by default, with no written backoff guidance (12 of 15). A Service Level Objective of 99.9 per cent per region on commercially reasonable efforts, dated 2 June 2026, and no SLA with remedies found (5 of 10). The docs carry no beta or preview label, though the specification's version field reads 0.0.0 (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": "security",
        "name": "Security \u0026 auth",
        "score": 85,
        "maxGain": 2.6,
        "reason": "API keys with a rotation period, expiry, regeneration and revocation, or OAuth 2.0 tokens from a service account. Keys are bound to one region and deployment profile and sent in a header (27 of 30). Strata Cloud Manager has a View Only Administrator role and custom roles with create, read, update and delete permissions per entity, and a scan key can't change configuration. Force-delete operations have no confirmation step (15 of 20). The service is itself an injection detector that scans tool inputs and outputs. Reports return the original payloads (13 of 15). An API scan log and log viewer in Strata Cloud Manager list each scan with its verdict, profile and action (15). PSIRT with a disclosure policy, a report form, a bug bounty and public advisories, and SOC 2+ on the certifications page without a product list we could read. No security.txt, and the onboarding sample curl passes `-k` (15 of 20).",
        "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": 71,
        "maxGain": 2.5,
        "reason": "Closed service under the EULA of August 2026, which names the contracting entity by region. The SDK is under the PolyForm Internal Use licence while the OpenAPI document's licence field says MIT (15 of 30). The privacy datasheet of 13 July 2026 lists the data processed, who can access it and retention (benign payloads up to 14 days, malicious up to 10 years), the EULA bars training generative AI on customer data, and a DPA is referenced. The 10-year period is disclosed, not hidden (26 of 30). No deprecation policy or dated notices for the API found, and the docs still mention a daily call quota alongside the monthly token model (0). GCP named as sub-processor with five countries, and out-of-region processing listed per detector (20).",
        "blend": "editorial 61, provenance 80",
        "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"
      }
    ],
    "provenance": [
      {
        "label": "Changelog",
        "value": "not found",
        "points": 0,
        "max": 10
      },
      {
        "label": "security.txt",
        "value": "not found",
        "points": 0,
        "max": 10
      }
    ],
    "unchecked": [
      "unchecked: the Platform Explorer page that lists new capabilities (https://docs.paloaltonetworks.com/platform-explorer), which is drawn by script, so dated notes for the API in the last 90 days may exist.",
      "unchecked: the MCP server's tool list and annotations, which need a key. The docs name `pan_inline_scan` and log a batch tool as `pan_batch_tool`.",
      "unchecked: commit dates through the GitHub API, which was rate-limited for our address. Dates come from a shallow clone of PaloAltoNetworks/pan.dev.",
      "Which certifications cover Prisma AIRS. The certifications page lists SOC 2+, FedRAMP, C5, IRAP and PCI DSS without a product list we could read, and the docs say the API intercept isn't available in FedRAMP environments.",
      "The money price of a billion tokens a month, and whether failed scans count against the quota.",
      "Whether the onboarding page's daily API call quota still applies to any customer, since the deployment profile page describes monthly tokens.",
      "The EULA forbids publishing benchmark or comparison tests and load testing, which matters before any probe of ours runs.",
      "`provenance.privacy` points at the product privacy datasheet, a PDF. The general Privacy Statement covers only data the company controls."
    ],
    "weaknesses": [
      "No public price, free tier or trial. Capacity is bought as Software NGFW credits, in steps of 1 billion tokens a month, through sales",
      "Payloads judged malicious are kept for up to 10 years, including after the subscription ends, per the privacy datasheet of 13 July 2026",
      "The EULA of August 2026 forbids publishing benchmark or comparison tests and forbids load testing of subscriptions",
      "Python is the only official SDK, and no dated changelog for the API was found",
      "Uptime is a 99.9 per cent objective on commercially reasonable efforts, with no SLA or credits found for Prisma AIRS"
    ],
    "agentNotes": [
      "Send `x-pan-token` and an `ai_profile` with `profile_name` or `profile_id` on every scan. Both come from Strata Cloud Manager, and a key works only in its own region",
      "Put earlier turns first in `contents`. Only the last element is scanned, and the rest is context",
      "Read `action` (allow or block) and `category`, then check `timeout` and `error`, because a detector that failed is reported in `errors` with a 200",
      "Keep synchronous requests under 2 MB and asynchronous ones under 5 MB and 25 items, and fetch at most 5 scan or report IDs a call, 10 calls a minute",
      "For MCP, connect to `/mcp` on the regional host with `x-pan-token` and `x-pan-profile` headers, and call `pan_inline_scan` yourself before and after generation"
    ],
    "recheck": "https://www.anchorterminal.com/builders/#disputes"
  },
  "meta": {
    "attribution": "Anchor Terminal (https://www.anchorterminal.com)",
    "docs": "https://www.anchorterminal.com/docs/",
    "generatedAt": "2026-10-09",
    "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"
  }
}