# Standing Framework

> Standing Framework builds the control layer for work that changes as it runs: consequential AI work, review records, proof boundaries, failure replay, and research artifacts.

Last updated: 2026-08-23
Canonical origin: https://standframe.com
Contact: info@standframe.com

## Primary Routes

- [Home](/) - Orientation page for Standing Framework and its primary routes: Products, Switchboard, Writing, and Research.
- [About](/about.html) - Company identity, category language, and proof boundary for Standing Framework.
- [Claims](/claims.html) - Registered public claim language, source state, proof boundary, approval requirements, and refresh rules.
- [Evidence-To-Decision Systems](/concepts/evidence-to-decision-systems.html) - Concept page connecting Standing Framework, Switchboard, essays, papers, and proof boundaries.
- [For Agents](/for-agents.html) - Task routing and crawler/content-use distinctions for answer engines and AI agents.
- [Products](/product.html) - Portfolio map organized by work state: Run, Review, Replay, and Improve.
- [Publication Status](/publications.html) - Lifecycle-aware publication status for website paper pages and blocked stronger publication claims.
- [Research](/research.html) - Research collaboration page for evidence, authority, recovery, review, workflow reconstruction, and proof packets.
- [Switchboard](/switchboard-control-plane.html) - Control surface for deciding what should happen next after consequential AI-agent work.
- [Switchboard Fit](/switchboard-fit.html) - Category and disambiguation page for Switchboard as a review/evidence-to-decision surface.
- [Writing](/writing.html) - Essay and research control surface for proof, authority, context, review, and failure notes.
- [Essays](/writing/essays/index.html) - Essay index for proof, authority, context, and reviewable agent work.
- [Research Papers](/writing/papers/index.html) - Research paper index with titles, authors, dates, and abstracts.

## Product Map

- Run: Switchboard is the clearest starting point for consequential AI work already in motion.
- Review: Signal Box is the review entry point for an AI-assisted software change.
- Replay: Test Stand is the failure-replay entry point for comparable scenarios.
- Improve: deeper system layers carry records, packages, extensions, and patterns forward.

## Research And Writing

- Essays: 6 public-safe essays on proof, authority, context, and reviewable agent work.
- Research papers: 31 abstract pages with authors, dates, and claim-bounded summaries.
- Use /writing/catalog.json when a structured index is more useful than prose.

## Machine Resources

- [Robots policy](/robots.txt) - Crawler access and sitemap discovery.
- [XML sitemap](/sitemap.xml) - Canonical route discovery for search and agent crawlers.
- [LLM orientation](/llms.txt) - Compact agent-readable map of the site, products, research, and claim boundaries.
- [Full LLM route map](/llms-full.txt) - Expanded agent-readable route inventory.
- [Writing catalog](/writing/catalog.json) - Structured essay and research-paper index.
- [OpenAPI description](/openapi.json) - OpenAPI 3.1 description for the read-only machine interface.
- [Machine interface home](/api/v1) - Compact source-backed API v1 home view for public agent discovery.
- [Machine interface status](/api/v1/status) - Local proof posture, held surfaces, and next actions for the API v1 surface.

## Claim Boundaries

- Local product proof is not customer adoption.
- Benchmark and research evidence is not product-market fit.
- Publication-ready website language is working positioning, not market validation.
- Production deployment and public promotion require separate authorization.
- Switchboard proof supports product and benchmark discipline, not general model superiority.
- The Control Console story is a product proof story, not external adoption evidence.
- Research pages are methods and evidence-bound notes unless a page states a separate publication status.
- Local corpus and manuscript evidence does not imply public representativeness.

## Crawler Policy

Crawler policy distinguishes search/index crawlers, user-requested fetch, and training/model-improvement crawlers. Production deployment still requires separate authorization.

- OAI-SearchBot: Allow / - Search/index discovery for ChatGPT search surfaces.
- ChatGPT-User: Allow / - User-requested fetch and browsing actions.
- GPTBot: Allow / - OpenAI training/model-improvement crawler access policy; this does not guarantee retrieval, training use, ranking, or citation.
- *: Allow / - General public crawler access for canonical public pages.