snapsynapse

Turnfile protocol execution guide for <AGENT_NAME>

Execute the Turnfile protocol (a SNAP protocol) in Claude for mailbox lifecycle, payload-first review, lock-safe shared-file updates, session start/close, Turnfile task/lock coordination, and OQ management. Use when the maintainer explicitly asks Claude to run a protocol module or produce an auditable protocol-conforming change.

Execute the Turnfile protocol (a SNAP protocol) in Google Antigravity for mailbox lifecycle, payload-first review, lock-safe shared-file updates, session start/close, Turnfile task/lock coordination, and OQ management. Use when the maintainer asks to run a protocol module, process the mailbox, do a payload-first review, or make an auditable coordination edit.

Execute the Turnfile protocol workflow in Codex for mailbox lifecycle, payload-first review exchange, lock-safe shared-file updates, session close/resume, and Turnfile task/lock coordination. Use when the maintainer explicitly asks Codex to run a protocol module or produce an auditable protocol-conforming change.

Execute the Turnfile protocol (a SNAP protocol) in Gemini CLI for mailbox lifecycle, payload-first review, lock-safe shared-file updates, session start/close, Turnfile task/lock coordination, and OQ management. Use when the maintainer explicitly asks Gemini to run a protocol module or produce an auditable protocol-conforming change.

Portable provenance, integrity, and drift control for Agent Skills bundles and their associated files across local folders, registries, platform uploads, and multi-agent sessions. Use when creating, editing, versioning, validating, packaging, or handing off a skill bundle; when checking or updating MANIFEST.yaml, CHANGELOG.md, hashes, stale evals, or frontmatter mode; and when keeping version identity with the bundle instead of filenames. Compatible with the agentskills.io open standard.

Execute the Turnfile protocol (a SNAP protocol) in Claude for mailbox lifecycle, payload-first review, lock-safe shared-file updates, session start/close, Turnfile task/lock coordination, and OQ management. Use when the maintainer explicitly asks Claude to run a protocol module or produce an auditable protocol-conforming change.

Execute the Turnfile protocol workflow in Codex for mailbox lifecycle, payload-first review exchange, lock-safe shared-file updates, session close/resume, and Turnfile task/lock coordination. Use when the maintainer explicitly asks Codex to run a protocol module or produce an auditable protocol-conforming change.

Triage and resolve a verification issue from the AI Capability Reference. Reads the GitHub issue, compares against current data, researches if needed, updates data files, and closes the issue with an audit comment.

Audit one or more capabilities for implementation coverage gaps. Reads capability definitions and the implementation index, cross-references against platform files, classifies each missing product as a mapping gap (existing feature not tagged) or data gap (feature not documented), presents findings, and applies fixes after confirmation.

Evaluate AI product announcements, model releases, pricing changes, and candidate providers against the AI Capability Reference project's current ontology and scope rules. Use when deciding whether something should be added, updated, watched, or skipped in the reference, especially for consumer-facing AI products, self-hosted runtimes, and open model access.

Draft the weekly AI Capability Reference roundup post for LinkedIn. Use this skill when it is time to write the Tuesday roundup, when the user says "draft the roundup," "Tuesday post," "what changed this week," or asks for a LinkedIn post summarizing recent capability reference updates. Also trigger when the user asks to log a published roundup or review the roundup cadence. This skill reads scan results, selects the right template, drafts a Post Protocol-compliant post, and logs publication to LocalBrain.

Portable provenance, integrity, and drift control for Agent Skills bundles and their associated files across local folders, registries, platform uploads, and multi-agent sessions. Use when creating, editing, versioning, validating, packaging, or handing off a skill bundle; when checking or updating MANIFEST.yaml, CHANGELOG.md, hashes, stale evals, or frontmatter mode; and when keeping version identity with the bundle instead of filenames. Compatible with the agentskills.io open standard.

Assess any API or website's Graceful Boundaries conformance level and provide concrete guidance for reaching the next level. Use this skill when the user asks to check a URL's rate limit communication, evaluate API conformance to Graceful Boundaries, assess how a service handles 429s, or improve how an API communicates its limits to agents. Also trigger when the user says "check this API's boundaries", "what level is this service", "assess graceful boundaries compliance", "how does this API handle rate limits", or provides a URL and asks about rate limit quality, conformance, or structured refusal. Distinct from the agent-readiness-audit skill, which assesses overall AI discoverability -- this skill specifically evaluates rate limit communication quality per the Graceful Boundaries specification.

Interactive guide for designing Knowledge-as-Code ontologies. Walks through domain analysis and entity role mapping to produce a draft project.yml. Use this skill when starting a new KaC project, when the user says "design an ontology", "set up a knowledge base", or wants to map a domain to the Primary/Container/Authority/Secondary entity model.

Run deterministic accessibility audits and regression gates on web projects, especially large sites with many routes and shared templates. Combines axe-core and optional Lighthouse with template-aware sampling, stable finding fingerprints, accepted baselines, standards evidence mapping (WCAG 2.1 AA default, WCAG 2.2 AA, EN 301 549), manual check guidance, and structured reporting. Output is configurable: markdown report only, markdown plus machine-readable JSON, or markdown plus issue tracker integration. Use this skill whenever the user mentions "accessibility audit", "a11y audit", "WCAG audit", "accessibility check", "compliance scan", "accessibility baseline", "new accessibility regressions", or asks to check a web project for accessibility issues. Also trigger when the user wants evidence for WCAG conformance review or mapping to a specific standard (WCAG 2.2, EN 301 549 / European Accessibility Act, CAN-ASC-6.2, ADA/AODA).

Use this skill whenever the user needs to generate, validate, or work with human-friendly unique identifiers. Trigger when the user mentions HardGuard25, unambiguous IDs, human-readable codes, or needs identifiers for order numbers, ticket codes, serial numbers, license keys, promo codes, booking references, tracking numbers, patient IDs, device IDs, short codes, or any scenario where an ID will be read, typed, printed, or spoken by a human. Also trigger when the user asks about ID alphabet design, character ambiguity, or compares encoding schemes like Crockford Base32.

Work with AI Decision Records (AIDR) in any repo with a decisions/ directory. Open a new record from the template, record an independent position, file an objection, assemble positions into a record, check lint status and earned claims, and request human arbitration. Use when the user says "open an AIDR", "record my position", "file an objection", "check the decision record", "request arbitration", or asks to document a consequential decision made with AI agents. NEVER authors arbitration.