Disclosure
What we claim, and what we do not.
A claim without its limitation is half the truth. This matrix pairs each thing AIQT does with what it does not do, and points every pairing at the evidence. Where a fact is not yet published, it points to where it will appear rather than implying it.
The matrix
Claim and limitation, side by side
What exists today
Claim: AIQT 1.0.5, the chat assistant, is released and downloadable today: a Skill for Claude, and a portable instruction file intended for any assistant that accepts standing instructions. The development assistant, version 1.1.0, is in development. Teams and Enterprise are ideas under consideration, with no dates and no commitments.
Limitation: No platform test result is published yet: the evidence page marks every platform pending, and ChatGPT, Gemini, and Copilot as not yet tested, so compatibility with a given assistant is intended, not verified. A 1.1.0 feature described on this site is a design, not a shipped product; it ships when it does what these pages say it does. Nothing labelled an idea is promised at all.
Evidence: The 1.0.5 release Where 1.1.0 stands
What the product is made of
Claim: The 1.0.5 artefacts are prose instructions and reference material only, with no executable code and no network calls.
Limitation: Instructions cannot do anything your assistant platform cannot already do. The 1.1.0 design adds local tooling whose exact behaviour is documented before it ships.
Evidence: What the files are The 1.1.0 directory
What it can enforce
Claim: AIQT is a behavioural standard your assistant is required to follow: it announces what its guardrails catch, surfaces issues instead of hiding them, and backs its claims with evidence.
Limitation: It does not technically prevent a model from erring. Results still depend on the model, the platform, competing instructions, and the tools in the conversation. AIQT does not make an AI infallible, and it does not sandbox, monitor, or block anything at runtime.
Evidence: What AIQT does not do
Security coverage
Claim: AIQT's source corpus carries a universal, language-neutral security baseline: behavioural rules covering secrets, authentication, authorization, input validation, untrusted content, logging, and data minimization. The 1.0.5 chat skill ships the subset a chat assistant can act on directly: keeping secrets out of the transcript, treating pasted or fetched content as data, resisting social pressure, and sending only the data a task needs. The fuller per-language and development-time delivery is a 1.1.0 design, not yet shipped.
Limitation: It is not a static analyzer, a vulnerability scanner, a penetration test, or an audit, and it does not guarantee that generated code is secure. Deeper per-language depth is a 1.1.0 design, composed at install from third-party sources you choose, which AIQT lists and credits but does not author or vouch for.
Evidence: Per-language depth, composed at install The frameworks the rules map to
Standards mappings
Claim: AIQT publishes a navigational crosswalk from the pack's rules to identifiers in security and AI-governance frameworks (OWASP, NIST, MITRE ATLAS, ISO/IEC, CSA), each pinned to a stated edition, with every mapped identifier validated against that edition before it can ship.
Limitation: A mapping asserts a relationship of ideas, not certification, attestation, audit evidence, or proof that you meet a standard. No framework publisher endorses, sponsors, or is affiliated with the pack. The absence of a mapping means none has been asserted yet, and a pinned edition may lag the newest release.
How AIQT itself is verified, and its change record
Claim: Substantive changes to the pack are reviewed before they land by independent verifiers from different model families, each briefed to refute rather than confirm. This project holds its own pack to a standing floor of three families; the pack's portable rule for adopters is two families for substantive work, and a third for critical changes. Every substantive change is recorded; the public change log is a curated release-level view.
Limitation: This describes how the AIQT project verifies its own pack; it says nothing about whether changes you make with AIQT installed receive the same review. Routine bookkeeping changes run the mechanical gates only.
Evidence: One QA standard, whichever model checks The change log
Data handling
Claim: The 1.0.5 pack sends nothing anywhere: it is instructions your assistant reads, with no code and no network calls. What you type still goes to your assistant's vendor under that vendor's terms, exactly as it would without AIQT. In the 1.1.0 design, routine checks run locally and send nothing to a model.
Limitation: The 1.1.0 substantive and delicate review tiers will send content to model providers. The providers, credentials, retention, and cost stay marked pending until they are documented before 1.1.0 ships. AIQT cannot intercept, encrypt, or block data that your assistant platform itself sends.
Platform support and currency
Claim: The evidence page carries the single platform test table: which platforms have a documented install method, when each was last tested, and the result. An untested platform says so plainly.
Limitation: As of today no platform test result is published; every last-tested entry is pending. Platform behaviour changes outside AIQT's control, so a test date states what was true that day, not a promise that it stays true. A per-file date on a rule says when the file last changed, not that it was re-verified on every platform that day.
Evidence: Platform test status
Licence, warranty, and endorsement
Claim: AIQT is open source under the Apache License 2.0 (except vendored third-party material, which remains under its own terms; see the NOTICE): you may use, copy, modify, distribute, and prepare derivative works of it freely, including commercially, and you keep your changes. If you redistribute it you keep the licence and copyright notices, carry prominent notices in any files you modify stating that you changed them (Apache License 2.0, section 4(b)), and retain the applicable attribution notices from the NOTICE (Apache License 2.0, section 4(d)).
Limitation: It is provided as-is, with no warranty or condition of any kind. No trademark rights are licensed (Apache License 2.0, section 6), and it carries the patent grant the Apache License 2.0 provides (Apache License 2.0, section 3). Using AIQT does not mean the project or its maintainer endorses you, and nothing permits you to imply sponsorship or official status. Equally, no framework publisher or platform vendor endorses AIQT.
Not legal or compliance advice
Claim: AIQT can sit alongside your compliance framework and help you demand more checkable behaviour from your assistant.
Limitation: It is not legal advice, not a compliance certification, and no substitute for your own counsel, policy, regulator, or production approval gates. Responsibility for what you ship stays with you.
Evidence: What AIQT does not do