Establish audited credibility baseline and founder approval gate

.github/workflows/founder-approval-gate.yml

@@ -0,0 +1,157 @@
+name: Founder Approval Gate
+
+on:
+  pull_request_target:
+    types: [opened, reopened, synchronize, edited, labeled, unlabeled]
+  issue_comment:
+    types: [created, edited, deleted]
+
+permissions:
+  contents: read
+  issues: write
+  pull-requests: write
+
+jobs:
+  founder-approval:
+    if: >-
+      github.event_name == 'pull_request_target' ||
+      (github.event_name == 'issue_comment' && github.event.issue.pull_request)
+    runs-on: ubuntu-latest
+    env:
+      FOUNDER_APPROVER_LOGIN: ${{ vars.FOUNDER_APPROVER_LOGIN }}
+
+    steps:
+      - name: Check governed paths and founder authorization
+        uses: actions/github-script@v7
+        with:
+          script: |
+            const owner = context.repo.owner;
+            const repo = context.repo.repo;
+            const pull_number = context.issue.number;
+            const founderLogin = process.env.FOUNDER_APPROVER_LOGIN || 'altmanAI';
+            const approvalPhrase = 'All Clear for Impact';
+            const marker = '<!-- altmanai-founder-approval-gate -->';
+
+            const { data: pull } = await github.rest.pulls.get({
+              owner,
+              repo,
+              pull_number,
+            });
+
+            const files = await github.paginate(
+              github.rest.pulls.listFiles,
+              { owner, repo, pull_number, per_page: 100 },
+            );
+
+            const governedRootFiles = new Set([
+              'GOVERNANCE.md',
+              'SECURITY.md',
+              'CONTRIBUTING.md',
+              'CODE_OF_CONDUCT.md',
+              'AI_SYSTEMS_POLICY.md',
+              'VERSION',
+              'CHANGELOG.md',
+              'package.json',
+              'pyproject.toml',
+              'Cargo.toml',
+              'go.mod',
+            ]);
+
+            const requiresFounderApproval = files.some(({ filename }) => {
+              return (
+                governedRootFiles.has(filename) ||
+                filename.startsWith('docs/') ||
+                filename.startsWith('profile/') ||
+                filename.startsWith('.github/workflows/') ||
+                filename.startsWith('.github/ISSUE_TEMPLATE/') ||
+                filename === '.github/pull_request_template.md' ||
+                /(^|\/)(release|version)(\.|\/|$)/i.test(filename)
+              );
+            });
+
+            if (!requiresFounderApproval) {
+              core.notice('No governed paths detected. Founder approval gate is not required.');
+              return;
+            }
+
+            const comments = await github.paginate(
+              github.rest.issues.listComments,
+              { owner, repo, issue_number: pull_number, per_page: 100 },
+            );
+
+            const exactPhrasePattern = /(?:^|\n)\s*All Clear for Impact\s*(?:\n|$)/;
+            const approvalComment = comments.find((comment) => {
+              return (
+                comment.user?.login === founderLogin &&
+                exactPhrasePattern.test(comment.body || '')
+              );
+            });
+
+            const changedGovernedFiles = files
+              .map(({ filename }) => filename)
+              .filter((filename) => {
+                return (
+                  governedRootFiles.has(filename) ||
+                  filename.startsWith('docs/') ||
+                  filename.startsWith('profile/') ||
+                  filename.startsWith('.github/workflows/') ||
+                  filename.startsWith('.github/ISSUE_TEMPLATE/') ||
+                  filename === '.github/pull_request_template.md' ||
+                  /(^|\/)(release|version)(\.|\/|$)/i.test(filename)
+                );
+              });
+
+            const status = approvalComment ? 'APPROVED' : 'PENDING';
+            const checklist = [
+              marker,
+              '## Founder approval gate',
+              '',
+              `**Status:** ${status}`,
+              `**Required approver:** @${founderLogin}`,
+              '',
+              'This pull request changes governance-sensitive, documentation, workflow, profile, release, or version-controlled paths.',
+              '',
+              '- [ ] Evidence and validation are linked in the pull request',
+              '- [ ] AI assistance is disclosed',
+              '- [ ] P.A.I.H.I. review is complete',
+              '- [ ] Rollback or correction plan is documented',
+              `- [${approvalComment ? 'x' : ' '}] Founder posted the exact standalone phrase: **${approvalPhrase}**`,
+              '',
+              '<details><summary>Governed files detected</summary>',
+              '',
+              ...changedGovernedFiles.map((filename) => `- \`${filename}\``),
+              '',
+              '</details>',
+              '',
+              approvalComment
+                ? `Approval record: ${approvalComment.html_url}`
+                : 'Approval must be a visible pull-request comment from the configured founder account. The workflow cannot create or infer approval.',
+            ].join('\n');
+
+            const priorGateComment = comments.find(
+              (comment) => comment.user?.type === 'Bot' && comment.body?.includes(marker),
+            );
+
+            if (priorGateComment) {
+              await github.rest.issues.updateComment({
+                owner,
+                repo,
+                comment_id: priorGateComment.id,
+                body: checklist,
+              });
+            } else {
+              await github.rest.issues.createComment({
+                owner,
+                repo,
+                issue_number: pull_number,
+                body: checklist,
+              });
+            }
+
+            if (!approvalComment) {
+              core.setFailed(
+                `Founder approval required. @${founderLogin} must post the exact standalone phrase: ${approvalPhrase}`,
+              );
+            } else {
+              core.info(`Founder approval verified at ${approvalComment.html_url}`);
+            }

CREDIBILITY_SCORECARD.md

@@ -0,0 +1,209 @@
+# AltmanAI Credibility Scorecard
+
+**Baseline audit date:** July 13, 2026  
+**Scope:** 14 public repositories currently visible under `github.com/altmanAI`  
+**Active compliance denominator:** 13 repositories  
+**Archived/historical:** 1 repository (`PAIHI-Portables`)  
+**Status:** Baseline assessment — not a certification or external audit
+
+> Credibility is not claimed. It is supported by inspectable evidence, limitations, decision records, and accountable human approval.
+
+## Executive finding
+
+AltmanAI's organization-level policy layer is materially stronger than its repository-level enforcement layer.
+
+The organization already publishes governance, security, contribution, conduct, AI-assistance, P.A.I.H.I., rollback, and founder-approval expectations. The primary credibility gap is that evidence remains distributed across README files, tests, manifests, pull requests, and registry artifacts instead of being indexed through consistent `GOVERNANCE.md`, `PROOF.md`, and `DECISION_LOG.md` files in each strategic repository.
+
+The active-portfolio baseline is **62.4/100** under the methodology below. This means the organization has a meaningful governance foundation, but it has not yet demonstrated consistent, automated, repo-level enforcement.
+
+## Scoring method
+
+| Dimension | Weight | What earns credit |
+|---|---:|---|
+| Governance adherence | 20 | Organization baseline, repo-specific authority, change classes, escalation and rollback rules |
+| Proof density | 25 | Tests, CI, reproducible commands, artifacts, citations, limitations and claim-to-evidence mapping |
+| P.A.I.H.I. implementation | 25 | Proof, Alignment, Integrity, Humanity and Impact demonstrated in the repository |
+| AI-assistance disclosure | 15 | Material PRs identify the system, assistance type, human verification and limitations |
+| Founder approval trail | 15 | Attributable approval record containing `All Clear for Impact` for material or governance-critical changes |
+
+Ratings:
+
+- **Strong:** 80–100
+- **Substantial:** 70–79
+- **Developing:** 50–69
+- **Foundational:** below 50
+- **Archived:** excluded from active compliance calculations
+
+Scores are conservative. Anything not directly verified during this baseline is treated as unverified rather than compliant.
+
+## Public repository scorecard
+
+| Repository | Positioning | Score | Rating | Verified evidence | Primary gap |
+|---|---|---:|---|---|---|
+| `.github` | Active governance layer | 86 | Strong | Organization governance, security, contribution rules, conduct rules, PR template, recent AI disclosures and founder approvals | Approval rules are not yet enforced automatically |
+| `canon-charter-v1` | Frozen canonical artifact | 82 | Strong | Repo governance, manifest, P.A.I.H.I. record, verification script, CI, changelog, version and approval phrase | No root `PROOF.md`; legacy AI label remains in a minted record |
+| `human-first-ai` | Working reference implementation | 79 | Substantial | Runnable architecture, tests, P.A.I.H.I. scorer, explicit limitations and founder approval | Material PR omitted explicit AI-system identity; no root `PROOF.md` |
+| `AltmanAI-Impact-Explorer` | Early full-stack implementation | 76 | Substantial | FastAPI/React structure, basic tests, public-data boundary and a disclosed correction PR | Dependency and deployment evidence not fully indexed; no root `PROOF.md` |
+| `altmanai-sdk` | Initial public foundation | 63 | Developing | Cross-language structure, schemas, hashing/verification utilities, test commands and status labels | Production adoption and parity are not independently demonstrated; no root `PROOF.md` |
+| `altmanai-master-ledger` | Active proof layer | 62 | Developing | Public ledger/registry structure, schemas, receipts and selected approval records | Evidence is partly self-attested and lacks one claim-to-artifact index |
+| `DailyPilot-Engine` | Working reference implementation | 62 | Developing | Runnable Python CLI, named modules, test command and explicit production limitation | CI/coverage/privacy/accessibility evidence is not indexed; no root `PROOF.md` |
+| `altmanai-masterplan` | Active execution map | 61 | Developing | Portfolio maturity labels, priorities, evidence-over-roadmap rule and founder approval requirement | No repo-specific governance, proof or decision log verified |
+| `altmanai-alt-seal` | Governance and verification asset | 57 | Developing | Published hashes, Save-Receipt artifacts, local verifier and scope limitations | Internally defined trust standard lacks independent validation and decision log |
+| `releaseshield-by-altmanai` | v0.1 product development | 50 | Developing | Clear scope, preliminary detection limitations, human review and proof-report design | Feature descriptions precede demonstrated working CLI/Action evidence |
+| `docs` | Active documentation | 45 | Foundational | Structured docs source, preview command and organization contribution routing | Published-site parity, accessibility and product-feature accuracy are not evidenced |
+| `paihi-framework` | Active development | 45 | Foundational | Development status, framework dimensions, human-authorization principles and reference implementation link | Organization-wide enforcement claims exceed currently demonstrated controls |
+| `altmanai-public-hub` | Active public entry point | 43 | Foundational | Clear company purpose, product direction, contacts and current/developing labels | Product and impact claims are weakly linked to code, tests and measurable outcomes |
+| `PAIHI-Portables` | Archived and consolidated | 40 | Archived | Honest archive notice states there was no working code and routes users to the SDK | Historical repository; excluded from active compliance denominator |
+
+## Governance adherence
+
+The `.github` repository supplies organization-level defaults for security reporting, contribution rules and conduct. It also declares its governance policy as the baseline for repositories without an approved local policy.
+
+However, GitHub community-health inheritance is not equivalent to a repo-specific authority map. Strategic repositories still need local rules for:
+
+- what counts as routine, material, security-sensitive and governance-critical;
+- who may approve each change class;
+- which files and release actions trigger founder approval;
+- incident escalation and rollback criteria;
+- the objective or masterplan record served by the repository.
+
+## Proof density
+
+Concrete proof is strongest in:
+
+1. `canon-charter-v1` — manifests, verification tooling and CI;
+2. `human-first-ai` — runnable reference code, tests and reproducible P.A.I.H.I. scoring;
+3. `altmanai-sdk` — schemas, cross-language foundations and deterministic artifact tooling;
+4. `DailyPilot-Engine` — working CLI, scoring modules and tests;
+5. `AltmanAI-Impact-Explorer` — full-stack structure and basic API tests.
+
+A root-level `PROOF.md` was not found in nine sampled strategic repositories: `altmanai-masterplan`, `altmanai-master-ledger`, `DailyPilot-Engine`, `altmanai-sdk`, `human-first-ai`, `releaseshield-by-altmanai`, `AltmanAI-Impact-Explorer`, `paihi-framework`, and `canon-charter-v1`.
+
+This does not mean those repositories have no evidence. It means the evidence is not yet normalized into a claim-to-proof index.
+
+## P.A.I.H.I. findings
+
+### Proof
+
+Tests, manifests, hashes and verification scripts exist in several repositories, but there is no organization-wide evidence threshold or conformance suite.
+
+### Alignment
+
+The current masterplan provides a clear portfolio map. Repo-specific links to formal objectives remain inconsistent.
+
+### Integrity
+
+Recent cleanup work has improved maturity labels and removed or corrected unsupported positioning. Remaining target-state language must stay distinguishable from shipped capability.
+
+### Humanity
+
+Human authorization is a documented principle. Accessibility, privacy, recourse and user-impact audits are not yet consistently evidenced per repository.
+
+### Impact
+
+Several repositories describe intended outcomes, but measurable success metrics and post-release observations are not consistently defined.
+
+## AI-assistance disclosure spot check
+
+The three-repository recent-PR spot check covered `.github`, `AltmanAI-Impact-Explorer`, and `human-first-ai`.
+
+| Repository | Material PRs reviewed | Result |
+|---|---:|---|
+| `.github` | 5 founder-authored material/governance PRs | Recent material PRs consistently identify AltmanAI Model 2.0 and human authorization; one cleanup PR requested approval but did not contain an explicit AI-system disclosure |
+| `AltmanAI-Impact-Explorer` | 1 material human-authored correction PR | AI assistance and `All Clear for Impact` are explicit |
+| `human-first-ai` | 1 material code/governance PR | Tests and founder approval are explicit; the assisting AI system is not named |
+
+**Confirmed correction item:** create a public correction record for the `human-first-ai` disclosure gap rather than rewriting history silently.
+
+Dependabot or automated dependency PRs should identify the automation source through GitHub metadata and are evaluated separately from human/AI-authored material changes.
+
+## Founder approval trail
+
+Visible approval records exist in recent governance, identity, canonical and material code changes. The phrase is currently a manual convention, not an enforced merge requirement.
+
+This baseline proposes a path-sensitive GitHub Action that requires an attributable comment containing exactly:
+
+> **All Clear for Impact**
+
+The workflow must not manufacture or automate founder approval.
+
+## Security and dependency posture
+
+### Public-safe finding
+
+A July 5, 2026 code-search sweep reported no exposed-secret hits, with an explicit warning that code-search indexing can lag.
+
+### Current limitation
+
+No new package-manager dependency audit, native GitHub secret-scanning verification, dependency-review run or third-party secret scan was completed as part of this connector-based baseline. Therefore:
+
+- this report does **not** claim zero vulnerabilities;
+- this report does **not** certify that no secret has ever been committed;
+- production repositories require live ecosystem-specific scans before the security success criteria can be checked.
+
+Sensitive findings, if discovered, must be reported privately under `SECURITY.md` and must not be added to this public scorecard before remediation.
+
+## Top 10 hardening scope
+
+The current strategic top 10 for the credibility initiative are:
+
+1. `.github`
+2. `altmanai-masterplan`
+3. `altmanai-master-ledger`
+4. `paihi-framework`
+5. `human-first-ai`
+6. `altmanai-sdk`
+7. `DailyPilot-Engine`
+8. `releaseshield-by-altmanai`
+9. `AltmanAI-Impact-Explorer`
+10. `canon-charter-v1`
+
+`PAIHI-Portables` is archived and is not treated as an active strategic implementation target.
+
+## Mandatory repository deliverables
+
+Each top-10 repository must add or explicitly link:
+
+- `GOVERNANCE.md`
+- `PROOF.md` or `EVIDENCE.md`
+- `DECISION_LOG.md`
+- dependency and supply-chain posture
+- secret-scanning posture
+- AI-assistance disclosure requirements
+- explicit success metrics and limitations
+- rollback and incident-escalation rules
+
+## Immediate correction queue
+
+1. Add founder-approval enforcement to the organization workflow.
+2. Create the `human-first-ai` AI-disclosure correction record.
+3. Add `PROOF.md` to the first implementation wave: `human-first-ai`, `altmanai-sdk`, `DailyPilot-Engine`, `ReleaseShield`, and `Impact-Explorer`.
+4. Recast target-state enforcement language in `paihi-framework` until conformance checks exist.
+5. Run dependency and secret scans and publish only remediated, public-safe summaries.
+6. Add measurable outcome definitions to product and public-facing repositories.
+
+## Monthly publication standard
+
+Future revisions should publish:
+
+- audit date and exact repository denominator;
+- score changes and methodology changes;
+- new evidence and validation links;
+- corrected or removed claims;
+- unresolved high-risk gaps;
+- dependency/security status at a public-safe level;
+- material PR disclosure compliance;
+- founder approvals for governance-critical changes.
+
+## Approval status
+
+This scorecard is proposed on a review branch.
+
+**Merge status:** Pending explicit Founder approval.  
+**Required phrase:** `All Clear for Impact`
+
+---
+
+**Owner:** AltmanAI Model 2.0 — audit execution and evidence scaffolding  
+**Accountable human authority:** Blake Hunter Altman, Founder & CEO  
+**Company:** AltmanAI by Altman Family Group LLC, Pittsburgh, Pennsylvania