Muhammad Atashi
Enterprise claims · evidence governancePUBLIC SOURCEV0.1.0RC2

TrustFlow

A claim-integrity workflow that binds generated answers to approved, current evidence; routes sensitive assertions through policy and human review; and tracks how source changes affect previously approved claims.

Owns
Binds every external claim to approved, current evidence
Status
Engineering release candidate
Stack
Python · Evidence registry · DOCX/XLSX · Review · Provenance
Public Reference

Hard problem

No evidence, no external claim.

Question · Can we substantiate this claim?

APPLIES TO

Security Teams · GRC · Compliance · Sales Engineering · Trust Centers · Enterprise Presales

PROOF

Evidence registry + claim provenance · policy/review gates · source-change impact analysis

CAPABILITIES

18 declared in the project record

  • Evidence registry
  • Claim provenance
  • Claim classification
  • Sensitivity rules
  • Evidence freshness
  • Conflict detection
  • Unsupported-claim blocking
  • Evidence-bounded generation
  • Review workflow
  • Safe document round-trip
  • Claim ledger
  • Source-change impact
  • Claim invalidation
  • Reusable approved-answer memory
  • Cross-questionnaire consistency
  • Coverage/reviewer analytics
  • Connector architecture
  • Export policy

Claim / evidence graph

Every external assertion keeps a lineage.

ASSERTION

Claim

classified assertion

EVIDENCE LINEAGE

No evidence, no external claim.

approved + fresh + scoped

01

Ledger

claim/evidence lineage

AUTHORITY

01

Policy

auto · review · block
02

Review

approve · edit · reject

Engineering decisions

The constraints that matter.

01

The claim is the governed object

The risk is an externally asserted organizational commitment, not merely a generated document.

02

Evidence eligibility includes freshness and scope

A once-correct source can become stale or be inapplicable to the claim being made.

03

Source changes propagate impact, not silent rewrites

Previously approved claims need explicit re-evaluation when their evidence changes.

Measurement contract

What would make the preferred answer fail?

EVALUATION METHOD

Automation is useful only when every external assertion stays traceable to approved, current, non-conflicting evidence.

Questionnaire fixture suites, intentionally stale/conflicting evidence, source-version mutation tests, reviewer replay, and export round-trips.

01

Unsupported claims

claims without valid evidence
02

Citation recall

supporting evidence surfaced
03

Conflict detection

contradictory evidence blocked
04

Staleness

expired evidence rejected
05

Impact recall

claims found after source change
06

Export fidelity

document round-trip integrity

Validation

What the evidence actually supports.

Claims stay scoped

Evidence registry

SUPPORTED

Owner · version · approval · effective/expiry dates · sensitivity · scope

Evidence metadata governs eligibility; it does not make the underlying source true by itself.

Claim provenance

SUPPORTED

Every asserted answer can retain exact supporting evidence lineage

Unsupported or conflicting evidence blocks automatic assertion.

Review policy

SUPPORTED

auto-eligible · review · block

Sensitive claim classes can require human review regardless of generated confidence.

Change impact

SUPPORTED

Source changes can identify dependent approved claims for re-review

Impact detection marks potential staleness; it does not silently rewrite historical answers.

Export boundary

SUPPORTED

Customer-facing output is separated from internal evidence payloads

Sensitive internal evidence is not automatically copied into external documents.

Boundaries

— Not generic RAG— Not a document chatbot— Not a full GRC suite— Not generic knowledge management
← Selected systemsNext · ResolveOps