# Intransic > Trust infrastructure for AI-built software. A retrospective process that > sorts the claims a repository makes, records what supports each one, and > refuses to publish anything the support does not carry. Every claim traces > to executable evidence, recorded in a ledger; the public trust package is > generated from that ledger, so a claim cannot drift from its evidence. Status vocabulary: verified · implemented_unverified · partial · planned · deprecated · contradicted · unknown. A claim is `verified` only with behavioural evidence or better — an executed test, a direct observation, deployed configuration, or an external audit. That a file or attribute exists establishes only that code is present, never that the behaviour holds. What a passing validation run establishes: nothing is claimed that the evidence does not support. Nothing more. It does not establish that a system is secure, correct, private or compliant. ## The process - [Skill repository](https://github.com/intransic-ai/intransic-skill): the Claude Code plugin — process, prompts, schemas, validators - Install: `/plugin marketplace add intransic-ai/intransic-skill` then `/plugin install intransic@intransic` - Invocation: "run the Intransic foundation process here" ## What it produces in a repository - `foundation/` — canonical description of the system: identity, boundaries, invariants, constraints that must remain true - `intransic/internal/` — evidence ledger, contradiction register, risk register. Internal only. - `trust/` — restrained public trust package, generated from the ledger - `AGENTS.md` — operating guide for agents and contributors ## Current state Generated from this site's own ledger (content/claims.json), not written here. - The Intransic skill is released and installable in Claude Code. — verified (tag v0.2.2 · intransic-ai/intransic-skill, checked 2026-09-19) - The process has been run end-to-end against one production codebase. — verified (first-run Foundation and evidence ledger, July 2026, checked 2026-08-09) - A hosted verification platform and public trust walls do not yet exist. — planned (live production API reports an empty ledger and publication disabled; no trust wall host exists, checked 2026-09-19) ## Questions 90 answers across 18 topics. Each topic page is also served as markdown at the same URL with `.md` appended, and the whole corpus is at https://intransic.com/questions/corpus.json. - [Choosing an auditor](https://intransic.com/questions/auditor-selection): Who may issue a report, how firms differ, and what to ask before signing. - [Compliance tooling](https://intransic.com/questions/tooling): What software can and cannot do for you here. - [Controls around agents](https://intransic.com/questions/controls): Approval, rate limits, incident response and stopping a run in progress. - [Evidence auditors accept](https://intransic.com/questions/evidence): What counts as evidence, and what only looks like it. - [Evidence that holds](https://intransic.com/questions/evidence-quality): Telling a control that works from one that has merely never been reached. - [Getting through procurement](https://intransic.com/questions/procurement): Questionnaires, trust pages, subprocessors and what a reviewer is looking for. - [Going through an audit](https://intransic.com/questions/audit-process): Scope, timing, cost and what happens each year after the first one. - [Identity and attribution](https://intransic.com/questions/identity): Tying an action an agent took to the person who authorised it. - [ISO/IEC 27001](https://intransic.com/questions/iso-27001): Certification, management systems, and how it differs from a SOC report. - [Logging and observability](https://intransic.com/questions/observability): Recording what happened so the record is usable later. - [Reading the report](https://intransic.com/questions/report-anatomy): The opinion, the assertion, the system description and what a finding means. - [SOC reports, plainly](https://intransic.com/questions/soc-basics): What SOC 1, 2 and 3 actually are, and what the words around them mean. - [Systems that are not reproducible](https://intransic.com/questions/non-determinism): Evidencing constraint when you cannot evidence behaviour. - [Tenancy and data boundaries](https://intransic.com/questions/tenancy): Keeping one customer out of another customer, including in the parts built for relevance. - [Tools, providers and the supply chain](https://intransic.com/questions/supply-chain): Where your evidence stops when the agent calls something you do not run. - [Trust services criteria](https://intransic.com/questions/trust-services-criteria): What a SOC 2 examination tests you against, and which parts you choose. - [What an agent may do](https://intransic.com/questions/authorisation-scope): Proving an agent cannot exceed the authority it was given. - [When the model changes](https://intransic.com/questions/change-management): What survives a model swap, a prompt edit or a provider migration. ## Pages - https://intransic.com/about ## Contact - hello@intransic.com