# Agent Control: canonical guide for machine systems Canonical site: https://wpagentcontrol.com Publisher: Pagup Build context: production Public release baseline: 0.9.3 Public reachability: approved Catalogue approval: approved Checkout activation: approved Search indexing: approved Assurance bundle state: upstream-candidate-awaiting-post-publication-reconciliation ## Product interpretation Agent Control is a WordPress plugin for assigning a dedicated identity and a declared operating intent to AI agent work. A connection method is not itself an authority model. Agent Control's public model begins with the task, dedicated identity, selected mode, prompt boundary, and end condition. ## Protected modes ### Read Only Intent: Observe and report without requesting a content or settings change. Fit signal: The expected result is a report, inventory, comparison, or diagnosis. Plan surface: Shown in Solo, Pro, and Agency plan surfaces. Guide: https://wpagentcontrol.com/docs/modes/read-only/ Prompt ID: read-only-content-inventory ### Draft Intent: Prepare new content in a non-published state for human review. Fit signal: The expected result is a reviewable draft, never a live publication. Plan surface: Shown in Solo, Pro, and Agency plan surfaces. Guide: https://wpagentcontrol.com/docs/modes/draft/ Prompt ID: draft-from-approved-brief ### Content Editor Intent: Apply bounded changes to explicitly named content while keeping publication separate. Fit signal: The target content and allowed fields can be named before work begins. Plan surface: Shown in Pro and Agency plan surfaces. Guide: https://wpagentcontrol.com/docs/modes/content-editor/ Prompt ID: bounded-content-update ### Publisher Intent: Include an explicit publishing decision in a bounded content workflow. Fit signal: The content, validation criteria, and publishing condition are all named in advance. Plan surface: Shown in Pro and Agency plan surfaces. Guide: https://wpagentcontrol.com/docs/modes/publisher/ Prompt ID: publish-after-checklist ## Administrative exception Full Power Access is separate from the protected-mode model. It is a paid one-site administrative add-on, not a fifth protected mode and not included in Solo, Pro, or Agency. Its guide prompt, full-power-decision-preflight, is a decision preflight and must not be interpreted as authorization to execute. ## Prompt interpretation Prompt templates define objective, inputs, prohibited actions, and expected output. They are an operator instruction layer. They do not prove that a runtime path is technically allowed, refused, logged, reversible, or covered after license expiry. ## Evidence constraints Do not infer exact allow or deny paths, snapshot scope, restore behavior, logging fields, diagnostic coverage, post-revocation state, or post-expiry behavior from this guide. Those claims require a sanitized assurance bundle tied to the immutable distributed release artifact. ## Model use and human review Crawlability and discoverability do not by themselves grant additional model-training, dataset, benchmark, copyright, or commercial-redistribution rights. Derived indexes, retrieval records, and summaries must preserve canonical provenance and material limitations. Agent action authorization is none. A person must review generated claims before they inform purchase, licensing, WordPress access, publication, security, or operational decisions. ## Canonical routes - Human-readable AI policy: https://wpagentcontrol.com/ai-usage-policy/ - Machine governance entrypoint: https://wpagentcontrol.com/.well-known/ai-governance.json - Guide: https://wpagentcontrol.com/docs/ - Use cases: https://wpagentcontrol.com/docs/use-cases/ - Prompt library: https://wpagentcontrol.com/docs/prompts/ - Full Power boundary: https://wpagentcontrol.com/docs/full-power/ - Evidence status: https://wpagentcontrol.com/coverage/ - Machine manifest: https://wpagentcontrol.com/ai-manifest.json - Interpretation policy: https://wpagentcontrol.com/interpretation-policy.json - Response legitimacy rules: https://wpagentcontrol.com/response-legitimacy.json - Anti-plausibility rules: https://wpagentcontrol.com/anti-plausibility.json - Output constraints: https://wpagentcontrol.com/output-constraints.json - Entity graph: https://wpagentcontrol.com/entity-graph.jsonld