Evidence review · technical
What is the miscsubjects build? — a formal external audit record
What is the miscsubjects build? — a formal external audit record
SEND THIS TO AN AUDITING MODEL https://miscsubjects.com/api/build-audit?format=drop
EXPLORE THE CAPABILITY INTERIOR https://miscsubjects.com/capability-atlas https://miscsubjects.com/api/capability-atlas
What is the miscsubjects build? — formal external-audit record
Generated: 2026-07-21T23:03:04.530Z
This is an evidence package, not a self-awarded conclusion. Every interpretation may be rejected by an external auditor.
Question
What is this build, end to end; where does it sit relative to current agent, workflow, protocol, and automation systems; what is it foremost at; and what should it be foremost at?
Provisional classification to test
Owner-operated, multi-model software and business operating environment with a public knowledge system and recursive development loop.
End-to-end system layers
Public knowledge and publishing system
Public articles, structured claims and sources, revisions, contributions, discourse, graph relationships, and model-facing representations.
Evidence: E5 — Public article index · E4 — Live cross-object graph · E10 — Source repository reference (authentication may be required)
Multi-model work environment
Coding agents, model-backed agents, resident agents, sessions, task queues, handoffs, and captured agent turns operate against shared build state.
Evidence: E1 — Live inventory of reachable files and objects · E6 — Configured model catalogue · E7 — Provider and gateway catalogue · E10 — Source repository reference (authentication may be required)
Capability and integration plane
Directory objects expose functions, HTTP integrations, agents, and flows through REST and MCP-compatible surfaces.
Evidence: E2 — Live self-describing capability registry · E3 — Whole-build route and binding map · E9 — Public object protocol article · E13 — Capability atlas joining current contracts, recorded invocations, tests, and coding-turn sediment
Cloud, local-computer, and external execution
The runtime binds cloud databases, object storage, queues, durable state, model execution, service workers, local computer operations, and external APIs.
Evidence: E3 — Whole-build route and binding map · E1 — Live inventory of reachable files and objects · E10 — Source repository reference (authentication may be required)
Business and communications operations
Messaging, marketing, commerce, analytics, customer, creative, and delivery capabilities coexist with software operations.
Evidence: E2 — Live self-describing capability registry · E1 — Live inventory of reachable files and objects
Evidence, history, and governance
Events, invocations, traces, rules, provenance, approvals, protected features, and turn records preserve operational history and constraints.
Evidence: E8 — Owner rule chain · E2 — Live self-describing capability registry · E10 — Source repository reference (authentication may be required)
Recursive software-development loop
Agents can inspect the repository and live state, edit implementation and data, deploy changes, and leave turn records for later agents.
Evidence: E1 — Live inventory of reachable files and objects · E10 — Source repository reference (authentication may be required) · E3 — Whole-build route and binding map · E13 — Capability atlas joining current contracts, recorded invocations, tests, and coding-turn sediment
Owner and public interfaces
Public articles, app surfaces, admin cockpit pages, graphs, ledgers, task views, and model handoffs expose different projections of the same build.
Evidence: E1 — Live inventory of reachable files and objects · E3 — Whole-build route and binding map · E5 — Public article index
Capability archaeology
The audit must inspect the accumulated capability interior, not only the public link surface.
json · 903 linestap to unfold
{
"summary": {
"registered_capabilities": 867,
"enabled_capabilities": 858,
"described_capabilities": 494,
"capabilities_with_recorded_invocations": 260,
"capabilities_with_registered_tests": 46,
"capabilities_with_currently_passed_tests": 1,
"registered_only": 591,
"recorded_invocations_across_current_capabilities": 143105,
"material_invocation_flags": 99044,
"categories": 108,
"domains": 12
},
"domains": [
{
"id": "cloud-data",
"name": "Cloud, data, storage, and deployment",
"registered": 133,
"enabled": 133,
"invoked": 29,
"with_registered_tests": 4,
"with_current_test_pass": 0,
"recorded_invocations": 5286,
"examples": [
{
"key": "BROWSER_JSON",
"description": "Extract LLM-structured JSON from a URL via Cloudflare Browser Rendering. $1=account_id, $2=JSON body {url, prompt?, response_format?}",
"verification": "registered_only"
},
{
"key": "BROWSER_LINKS",
"description": "Extract all links from a URL via Cloudflare Browser Rendering. $1=account_id, $2=JSON body {url}",
"verification": "invoked_not_test_proven"
},
{
"key": "BROWSER_MARKDOWN",
"description": "Get the markdown of a URL via Cloudflare Browser Rendering. $1=account_id, $2=JSON body {url}. Returns the rendered markdown",
"verification": "invoked_not_test_proven"
},
{
"key": "BROWSER_PDF",
"description": "Render a URL as PDF via Cloudflare Browser Rendering. $1=account_id, $2=JSON body {url}. Returns binary PDF",
"verification": "invoked_not_test_proven"
},
{
"key": "BROWSER_SCRAPE",
"description": "Extract structured data by selectors via Cloudflare Browser Rendering. $1=account_id, $2=JSON body {url, elements:[{selector}]}",
"verification": "invoked_not_test_proven"
},
{
"key": "BROWSER_SCREENSHOT",
"description": "Get a PNG screenshot of a URL via Cloudflare Browser Rendering. $1=account_id, $2=JSON body {url, screenshotOptions?}. Returns binary PNG",
"verification": "invoked_not_test_proven"
},
{
"key": "CF",
"description": "Cloudflare REST API unified entrypoint. 256+ operations.",
"verification": "invoked_not_test_proven"
},
{
"key": "CF_API_GAPS",
"description": "Analyze the CF row target_map. Returns counts, known namespaces, expected namespaces, missing namespaces, coverage_pct",
"verification": "registered_only"
},
{
"key": "CF_AUDITLOGS_AUDITLOGS_BY_ACCOUNT_ID",
"description": "Find all audit logs (a list of who made what change when) for a Cloudflare Account by ID. This can be used to query activity on your Cloudflare account at a particular time. Since and before are requi",
"verification": "registered_only"
},
{
"key": "CF_AUTORAG_LIST_RAGS",
"description": "List AutoRAGs (vector stores)",
"verification": "registered_only"
},
{
"key": "CF_AUTORAG_SEARCH",
"description": "Search Documents using AutoRAG (vector store)",
"verification": "registered_only"
},
{
"key": "CF_BINDINGS_D1_DATABASES_LIST",
"description": "List all of the D1 databases in your Cloudflare account",
"verification": "invoked_not_test_proven"
}
]
},
{
"id": "communications",
"name": "Messaging, email, phone, and voice",
"registered": 117,
"enabled": 115,
"invoked": 22,
"with_registered_tests": 4,
"with_current_test_pass": 0,
"recorded_invocations": 1710,
"examples": [
{
"key": "AUDIO",
"description": "Speak words aloud. INVOKE: [invocation omitted]",
"verification": "invoked_not_test_proven"
},
{
"key": "BLOCK_IMESSAGE",
"description": "BLOCK_IMESSAGE — iMessage layout (optional, not default)",
"verification": "disabled"
},
{
"key": "BLOCK_VOICE",
"description": "text",
"verification": "disabled"
},
{
"key": "BLOOIO",
"description": "Blooio (iMessage/SMS) unified entrypoint",
"verification": "invoked_not_test_proven"
},
{
"key": "BLOOIO2",
"description": "You are Cyrus's writing assistant on the second Blooio line (+12065711028). When he texts you, you write and edit articles on the miscsubjects build.",
"verification": "registered_only"
},
{
"key": "BLOOIO_ADD_CONTACT_IDENTITY",
"description": "Attach a new identity (phone/email on a channel type) to an existing contact.",
"verification": "registered_only"
},
{
"key": "BLOOIO_ADD_CONTACT_TAGS",
"description": "Add one or more tags to a contact (deduped).",
"verification": "registered_only"
},
{
"key": "BLOOIO_ADD_REACTION",
"description": "React to a message (tapback). reaction like '+love', '+like', '+laugh', '+emphasize', '+question', or remove with a '-' prefix. Blooio channels only.",
"verification": "registered_only"
},
{
"key": "BLOOIO_BATCH_LOOKUP_PHONE_NUMBERS",
"description": "Look up up to 100 phone numbers at once. Requires an enterprise plan.",
"verification": "registered_only"
},
{
"key": "BLOOIO_CONTACT_CAPABILITIES",
"description": "Check if a phone number or email is reachable via iMessage / SMS / RCS. Resolves (or creates) the contact for that identifier, then returns its capabilities. Note: creating the contact is a side effec",
"verification": "registered_only"
},
{
"key": "BLOOIO_CREATE_CHAT",
"description": "Find or create a 1:1 chat with a recipient on a channel. Returns the chat id (chat_...) to use for sends and actions.",
"verification": "registered_only"
},
{
"key": "BLOOIO_CREATE_CONTACT",
"description": "Create a contact, optionally with one identity (phone/email on a channel type). Returns the new contact id (ct_...).",
"verification": "registered_only"
}
]
},
{
"id": "computer-code",
"name": "Computer, code, files, browser, and GitHub",
"registered": 111,
"enabled": 111,
"invoked": 37,
"with_registered_tests": 1,
"with_current_test_pass": 0,
"recorded_invocations": 614,
"examples": [
{
"key": "BROWSER_FETCH",
"description": "{\"cmd\":\"sh\",\"args\":[\"-lc\",\"curl -sSL \\\"$1\\\" | head -c 20000\"],\"timeout\":60000}",
"verification": "invoked_not_test_proven"
},
{
"key": "BROWSER_PLAYWRIGHT",
"description": "Run a Playwright CLI command on the Mac (npx -y playwright …). g. \"screenshot https://x.com /tmp/x.png\". First run downloads browsers (~2 min). For full browser automation absorb the playwright MCP server via MCP_PROBE",
"verification": "invoked_not_test_proven"
},
{
"key": "BROWSER_USE",
"description": "Run the browser-use agent (AI browser driver). Installed at bridge/.venv-browser-use (Python 3.12). Needs GROK_API_KEY, OPENAI_API_KEY, or ANTHROPIC_API_KEY in ~/.config/grok-bridge.env",
"verification": "invoked_not_test_proven"
},
{
"key": "CLI_AIDER",
"description": "Aider — git-aware AI pair programmer running one-shot on the Mac.",
"verification": "registered_only"
},
{
"key": "CLI_AWS",
"description": "run aws CLI. ARGS: $1 = args. {\"cmd\":\"sh\",\"args\":[\"-lc\",\"aws $1+\"],\"timeout\":120000}",
"verification": "registered_only"
},
{
"key": "CLI_BREW",
"description": "run brew. ARGS: $1 = args. {\"cmd\":\"sh\",\"args\":[\"-lc\",\"brew $1+\"],\"timeout\":120000}",
"verification": "registered_only"
},
{
"key": "CLI_BUN",
"description": "run bun. ARGS: $1 = args, $2 = cwd. {\"cmd\":\"sh\",\"args\":[\"-lc\",\"cd \\\"$2\\\" 2>/dev/null; bun $1\"],\"timeout\":120000}",
"verification": "registered_only"
},
{
"key": "CLI_CLASP",
"description": "run clasp (Apps Script CLI) on Cyrus's Mac. ARGS: $1 = args after `clasp`. cwd = /Users/cyrusmassoumi/miscsubjects-pages/apps-script.",
"verification": "registered_only"
},
{
"key": "CLI_CLAUDE_CODE",
"description": "Claude Code headless coding agent on the Mac. File edit + shell.",
"verification": "invoked_not_test_proven"
},
{
"key": "CLI_CODEX",
"description": "OpenAI Codex CLI one-shot. Parallel/second-opinion coder.",
"verification": "registered_only"
},
{
"key": "CLI_CURL_LOCAL",
"description": "run curl ON THE MAC (different from the worker's outbound fetch — picks up Mac creds in keychain/env).",
"verification": "registered_only"
},
{
"key": "CLI_DENO",
"description": "run deno. ARGS: $1 = args, $2 = cwd. {\"cmd\":\"sh\",\"args\":[\"-lc\",\"cd \\\"$2\\\" 2>/dev/null; deno $1\"],\"timeout\":120000}",
"verification": "registered_only"
}
]
},
{
"id": "knowledge-content",
"name": "Knowledge, articles, graph, and protocol",
"registered": 88,
"enabled": 88,
"invoked": 46,
"with_registered_tests": 1,
"with_current_test_pass": 0,
"recorded_invocations": 67544,
"examples": [
{
"key": "ARTICLES",
"description": "Fast natural-language article CRUD. One row per article in D1; writes are URL-invocable through OIP.",
"verification": "invoked_not_test_proven"
},
{
"key": "ARTICLE_ASK",
"description": "Answer from article ledger topology (claims, sources, Reddit/X anecdotes, user reports). Not medical advice.",
"verification": "invoked_not_test_proven"
},
{
"key": "ARTICLE_CLAIM",
"description": "Post one tiered claim voxel into an article ledger with who_claims + posted_by provenance. Not medical advice.",
"verification": "invoked_not_test_proven"
},
{
"key": "ARTICLE_GET",
"description": "Return article + its slots as one joined result set. $1=slug. Returns an array of {slug,title,subject,updated_at,slot_key,content,model,version}.",
"verification": "invoked_not_test_proven"
},
{
"key": "ARTICLE_INGEST",
"description": "Parse user-submitted evidence and write to article source ledger + optional claims. Hash-chained.",
"verification": "invoked_not_test_proven"
},
{
"key": "ARTICLE_PUT",
"description": "create/replace an article. ARGS: $1 = JSON {slug,title,body,hero?,images?,style?} $1",
"verification": "invoked_not_test_proven"
},
{
"key": "ART_GET",
"description": "Read an article by slug",
"verification": "registered_only"
},
{
"key": "ART_PATCH",
"description": "Edit an article by slug",
"verification": "invoked_not_test_proven"
},
{
"key": "ARXIV_GROW",
"description": "Regenerate the arXiv paper from live state. Reads paper/template.tex + paper/rings.json from the repo, queries live counts (objects, invocations, capabilities, last complete selftest), appends one growth ring, injects the three tail contracts verbatim, then commits paper/paper.tex + paper/rings.json + README.md + oip.json — each commit message carries this trace id. CI compiles the PDF on the paper.tex push. This fn is the only writer of the generated files.",
"verification": "invoked_not_test_proven"
},
{
"key": "ARXIV_PAPER",
"description": "The arXiv paper as a live object. The paper \"The Document Is the Receipt\" lives at github.com/massoumicyrus/oip (private) and is written only by ARXIV_GROW. Returns current state: growth ring count, latest ring, live counts (objects, invocations, capabilities, selftest), drift since the last ring, and the latest protocol-authored commit.",
"verification": "invoked_not_test_proven"
},
{
"key": "CAPABILITY_ATLAS",
"description": "Read the public capability archaeology atlas joining every current directory contract with recorded invocation evidence, registered tests, capability domains, and aggregate coding-agent turn/file-change sediment. It separates registered, invoked, tested, and disabled states so the build interior can be audited without treating row count as proof.",
"verification": "registered_only"
},
{
"key": "CAP_EXPLAIN",
"description": "Explain a capability: what it may invoke, verbs, expiry + remaining TTL, uses left, risk ceiling, owner gate, revocation, ledger trail. Accepts the token itself (sh.…) or its fingerprint (cap_…). Never echoes the raw token.",
"verification": "invoked_not_test_proven"
}
]
},
{
"id": "growth",
"name": "Marketing, leads, sales, and distribution",
"registered": 82,
"enabled": 81,
"invoked": 35,
"with_registered_tests": 11,
"with_current_test_pass": 0,
"recorded_invocations": 2258,
"examples": [
{
"key": "ARCADS",
"description": "{{SHARED}}",
"verification": "invoked_not_test_proven"
},
{
"key": "ARCADS_CREDITS",
"description": "ArcAds credit usage this month. Returns {month,used,cap,remaining}. Cap from settings.arcads_monthly_credits (80440). Logged from each generate (data.creditsCharged)",
"verification": "invoked_not_test_proven"
},
{
"key": "ARCADS_GENERATE",
"description": "Generate an ad image via ArcAds, poll to completion, store to R2, return a stable link",
"verification": "invoked_not_test_proven"
},
{
"key": "ARCADS_ROUTES",
"description": "ArcAds HTTP unified entrypoint",
"verification": "invoked_not_test_proven"
},
{
"key": "ARCADS_TO_R2",
"description": "Poll a finished ArcAds render by arcads_id and re-store its bytes to R2, returning a stable https://miscsubjects.com/img/ link (permanent, not a presigned/expiring S3 url). Waits up to 120s for the render.",
"verification": "invoked_not_test_proven"
},
{
"key": "ARCADS_UPLOAD",
"description": "Upload a file to ArcAds (presign + S3 PUT). Returns {filePath,fileId}; pass filePath in referenceImages. fileType e.g. image/png, image/jpeg, video/mp4, audio/mp3",
"verification": "registered_only"
},
{
"key": "ARCADS_VIDEO_GENERATE",
"description": "Generate a video via ArcAds, poll, store to R2, return a stable link",
"verification": "invoked_not_test_proven"
},
{
"key": "KLAVIYO",
"description": "Klaviyo unified entrypoint",
"verification": "invoked_not_test_proven"
},
{
"key": "KLAVIYO_PROFILES",
"description": "DISABLED: stale provider stub. Use [invocation omitted] and report the returned page count/next-page state.",
"verification": "disabled"
},
{
"key": "LEADS_DISCOVER",
"description": "Find businesses likely to buy peptides wholesale or want white-label, by type and city. Free source (OpenStreetMap), no API key. Writes new rows to the leads table.",
"verification": "invoked_not_test_proven"
},
{
"key": "LEADS_DISCOVER_AI",
"description": "SCRAPER 2 — Grok live web search finds real businesses (with verified official websites) that OpenStreetMap does not have. Inserts them as new leads, source=grok-live-search.",
"verification": "invoked_not_test_proven"
},
{
"key": "LEADS_DISCOVER_PLACES",
"description": "Discover B2B leads via Google Places API (New) — med-spas/clinics/longevity by segment + city, with website/phone/rating. Better coverage than the Overpass source. Inserts into the leads table.",
"verification": "invoked_not_test_proven"
}
]
},
{
"id": "other",
"name": "Other and uncategorized",
"registered": 82,
"enabled": 77,
"invoked": 16,
"with_registered_tests": 6,
"with_current_test_pass": 0,
"recorded_invocations": 3611,
"examples": [
{
"key": "AGENT",
"description": "Control a resident agent",
"verification": "registered_only"
},
{
"key": "AIG_LIST",
"description": "List AI Gateways on the account",
"verification": "registered_only"
},
{
"key": "AIG_RAW",
"description": "Raw call to AI Gateway REST. Arg $1 = full JSON body (model + messages)",
"verification": "registered_only"
},
{
"key": "APPS_SCRIPT_RUN",
"description": "Generic Apps Script invocation. $1=action name, $2=JSON args. Routes to the airunner web app deployment which has Google Drive / Sheets / Tasks / Calendar / Apps Script-runtime access. Use when you need to read/write the user's Google stuff. $$2 (double dollar) inlines the args JSON raw, so callers pass {\"k\":\"v\"} not as a string but as a JSON value",
"verification": "invoked_not_test_proven"
},
{
"key": "ASK_CLAUDE",
"description": "{{SHARED}}",
"verification": "invoked_not_test_proven"
},
{
"key": "ASK_GEMINI",
"description": "{{SHARED}}",
"verification": "invoked_not_test_proven"
},
{
"key": "ASK_GPT",
"description": "{{SHARED}}",
"verification": "invoked_not_test_proven"
},
{
"key": "ASK_KIMI",
"description": "{{SHARED}}",
"verification": "invoked_not_test_proven"
},
{
"key": "BLOCK_EMOJI",
"description": "BLOCK_EMOJI — tapback / emoji reaction language",
"verification": "disabled"
},
{
"key": "BLOCK_REASONING_A",
"description": "BLOCK_REASONING_A — reasoning format (type A)",
"verification": "disabled"
},
{
"key": "BLOCK_ROUTING",
"description": "BLOCK_ROUTING — identity routing (ROUTER)",
"verification": "disabled"
},
{
"key": "CC_LAST",
"description": "the last 3 Claude Code turns on this repo (the user input to Claude Code, the tools it used, the files it edited).",
"verification": "registered_only"
}
]
},
{
"id": "commerce",
"name": "Commerce, payments, and customers",
"registered": 60,
"enabled": 60,
"invoked": 7,
"with_registered_tests": 4,
"with_current_test_pass": 0,
"recorded_invocations": 23,
"examples": [
{
"key": "BC",
"description": "BigCommerce unified entrypoint",
"verification": "invoked_not_test_proven"
},
{
"key": "CUSTOMER",
"description": "You are the customer-facing agent for miscsubjects.com. You answer people who reply to iMessage/Facebook ads or visit the site.",
"verification": "registered_only"
},
{
"key": "PAYMENTS_CREATE_REFUND",
"description": "This tool will refund a payment intent in Stripe. It takes three arguments: - payment_intent (str): The ID of the payment intent to refund. - amount (int, optional): The amount to refund in currency m",
"verification": "registered_only"
},
{
"key": "PAYMENTS_FETCH_STRIPE_RESOURCES",
"description": "Retrieve Stripe object details by ID. IMPORTANT: Only call this tool after search_stripe_resources is called to get specific object IDs. Do not use this tool to discover or search for objects. This to",
"verification": "registered_only"
},
{
"key": "PAYMENTS_GET_STRIPE_ACCOUNT_INFO",
"description": "This will get the account info for the logged in Stripe account.",
"verification": "registered_only"
},
{
"key": "PAYMENTS_SEARCH_STRIPE_DOCUMENTATION",
"description": "Search the Stripe documentation for the given question and language. It takes two arguments: - question (str): The user question to search an answer for in the Stripe documentation. - language (str, o",
"verification": "registered_only"
},
{
"key": "PAYMENTS_SEARCH_STRIPE_RESOURCES",
"description": "This tool can be used to search for specific Stripe resources using a custom Stripe query syntax. It is only able to search for the following resources: customers, payment_intents, charges, invoices,",
"verification": "registered_only"
},
{
"key": "PAYMENTS_STRIPE_API_DETAILS",
"description": "Get detailed parameter information for a specific Stripe API operation. Provide the stripe_api_operation_id from stripe_api_search results to see all path, query, and body parameters with their types,",
"verification": "registered_only"
},
{
"key": "PAYMENTS_STRIPE_API_READ",
"description": "Read data from any Stripe API GET operation: 1. Use stripe_api_search to find the operation ID. 2. Use stripe_api_details to understand its parameters (required for operations with nested object field",
"verification": "registered_only"
},
{
"key": "PAYMENTS_STRIPE_API_SEARCH",
"description": "Search for Stripe API operations by keyword. Returns matching operations with their HTTP method, path, summary, and top-level parameter names with types. Does not include parameter descriptions, enum",
"verification": "registered_only"
},
{
"key": "PAYMENTS_STRIPE_API_WRITE",
"description": "Write data via any Stripe API POST/PATCH/PUT/DELETE operation: 1. Use stripe_api_search to find the operation ID. 2. Use stripe_api_details to understand its parameters (required for operations with n",
"verification": "registered_only"
},
{
"key": "PAYMENTS_STRIPE_IMPLEMENTATION_PLANNER",
"description": "Stripe payment integration planner. Use this tool to help users accept payments, sell products online, set up billing, or build any Stripe integration. Call this BEFORE writing code when the user want",
"verification": "registered_only"
}
]
},
{
"id": "automation-work",
"name": "Automation, flows, tasks, and durable work",
"registered": 52,
"enabled": 52,
"invoked": 17,
"with_registered_tests": 5,
"with_current_test_pass": 0,
"recorded_invocations": 60944,
"examples": [
{
"key": "ADDTASK",
"description": "Add a job to the build queue (tasks table). Cron will pick it up and run it automatically when protocol_autorun is on.",
"verification": "invoked_not_test_proven"
},
{
"key": "AUTOMATE_ADD",
"description": "Automate a capability — make it fire itself every N minutes and ledger a receipt each run. Turns a one-off action into a standing background job.",
"verification": "invoked_not_test_proven"
},
{
"key": "AUTOMATE_DELETE",
"description": "Delete an automation by id.",
"verification": "invoked_not_test_proven"
},
{
"key": "AUTOMATE_FIRE",
"description": "Fire every enabled automation registered for an event (trigger=event:NAME). Any inbound hook (a customer text, an error, a webhook) calls this with the event name + payload; matching automations run and ledger a receipt.",
"verification": "invoked_not_test_proven"
},
{
"key": "AUTOMATE_LIST",
"description": "List every automation — id, name, schedule, target key, enabled, last run, run count.",
"verification": "invoked_not_test_proven"
},
{
"key": "AUTOMATE_RUN_DUE",
"description": "Fire every enabled automation whose interval has elapsed; ledger a receipt for each. Called by the cron each tick; safe to call manually.",
"verification": "invoked_not_test_proven"
},
{
"key": "AUTOMATE_TOGGLE",
"description": "Turn an automation on or off by id (paused, not deleted).",
"verification": "invoked_not_test_proven"
},
{
"key": "CRON_LIST",
"description": "List existing cron jobs by querying the Cloudflare API. Returns scheduled jobs for this account.",
"verification": "registered_only"
},
{
"key": "DELTASK",
"description": "delete a task. ARGS: <id>. EX {\"key\":\"DELTASK\",\"body\":\"42\"} [\"DELETE FROM tasks WHERE id=?\",\"$1\"]",
"verification": "registered_only"
},
{
"key": "DIRECTORY_GET",
"description": "Read one directory row by $1=key. Use to inspect a specific tool definition.",
"verification": "invoked_not_test_proven"
},
{
"key": "DIRECTORY_LIST",
"description": "List the directory: D1_QUERY: SELECT key, type, target, updated_at FROM directory ORDER BY key. Use to enumerate every tool.",
"verification": "invoked_not_test_proven"
},
{
"key": "DIRECT_EXEC_IF_PREFIX",
"description": "If the inbound message starts with /t /exec /terminal /run /help, treat the rest as a LOCAL_EXEC command on the Mac bridge. Otherwise pass through unchanged. $1=full message.",
"verification": "registered_only"
}
]
},
{
"id": "agents-models",
"name": "Agents and model runtimes",
"registered": 50,
"enabled": 50,
"invoked": 20,
"with_registered_tests": 1,
"with_current_test_pass": 0,
"recorded_invocations": 288,
"examples": [
{
"key": "AGENT_BRIDGE",
"description": "Route a message from one agent to another. $1=source_agent, $2=target_agent, $3=message. Creates a turn that appears to come from the source agent, routed to the target agent.",
"verification": "registered_only"
},
{
"key": "AGENT_IMPORT",
"description": "Cannibalize an agent definition (md frontmatter name/description/model/tools) into a proposed agent row. PROPOSE only",
"verification": "registered_only"
},
{
"key": "AGENT_LEARN",
"description": "save a lesson/working pattern to an agent persistent memory. Args: agent|lesson.",
"verification": "registered_only"
},
{
"key": "AGENT_LIST",
"description": "List resident agents and their live status",
"verification": "invoked_not_test_proven"
},
{
"key": "AGENT_RECALL",
"description": "load an agent saved lessons/patterns. Arg: agent.",
"verification": "registered_only"
},
{
"key": "AGENT_SPAWN",
"description": "Spawn a resident agent that loops on a goal until done (durable, survives Mac sleep)",
"verification": "registered_only"
},
{
"key": "AGENT_SPAWN_CLI",
"description": "Alias of CLI_SPAWN for router tags. Args: agent|prompt|cwd|mode|delivery",
"verification": "registered_only"
},
{
"key": "AGENT_TURNS",
"description": "last N agent turns across all CLI agents (claude, codex, grok, …). $1 = agent id or \"all\", $2 = limit (default 5).",
"verification": "invoked_not_test_proven"
},
{
"key": "AGENT_TURNS_FILTER",
"description": "multi-filter agent turn query. $1=agent|all, $2=tag|risk|all, $3=limit.",
"verification": "registered_only"
},
{
"key": "AGENT_TURNS_ISSUES",
"description": "agent turns matching issue tags (risk, protected, file_edit, unaudited, audit_fail). $1 = tag, $2 = limit.",
"verification": "registered_only"
},
{
"key": "AGENT_TURNS_TRACE",
"description": "agent turns linked to one ledger trace_id. $1 = trace_id, $2 = limit (default 20).",
"verification": "registered_only"
},
{
"key": "CF_AI_GATEWAY_GET_LOG_DETAILS",
"description": "Get a single Log details",
"verification": "registered_only"
}
]
},
{
"id": "system-meta",
"name": "System, directory, settings, and self-modification",
"registered": 50,
"enabled": 50,
"invoked": 22,
"with_registered_tests": 10,
"with_current_test_pass": 1,
"recorded_invocations": 655,
"examples": [
{
"key": "ADD_ROW",
"description": "Insert a new row into the directory (tools, agents, flows).",
"verification": "invoked_not_test_proven"
},
{
"key": "BUILDER",
"description": "{{SHARED}}",
"verification": "registered_only"
},
{
"key": "BUILDER_ADD",
"description": "append a row to builder_queue. status defaults to \"idea\", priority to 5.",
"verification": "invoked_not_test_proven"
},
{
"key": "BUILDER_DELETE",
"description": "delete a row from builder_queue. ARGS: id. [\"$1\"]",
"verification": "invoked_not_test_proven"
},
{
"key": "BUILDER_DONE",
"description": "mark id done + record proof (link/exhibit). ARGS: id|proof.",
"verification": "registered_only"
},
{
"key": "BUILDER_LIST",
"description": "return the current builder queue as JSON, sorted by priority. ARGS: optional status filter (idea|queued|in_progress|blocked|done|wont).",
"verification": "invoked_not_test_proven"
},
{
"key": "BUILDER_NEXT",
"description": "pick the single highest-priority queued item.",
"verification": "registered_only"
},
{
"key": "BUILDER_PATCH",
"description": "update one row's status/priority/title/body/blocker/proof. ARGS: id|field|value.",
"verification": "registered_only"
},
{
"key": "CARDS",
"description": "Read the STATE CARDS. One card per isolated event: the inbound message (plain text), the reply (plain text), the agent system prompt, and every tool/CLI/file/http step with its raw payload in/out — each classified. Every card has a trace id (card_id) and a content hash, and is REST-callable",
"verification": "registered_only"
},
{
"key": "CATEGORIES",
"description": "List all tool categories.",
"verification": "invoked_not_test_proven"
},
{
"key": "DEDUP_INSERT",
"description": "Insert message_id into blooio_dedup. Use ONLY from the inbound webhook handler before replying",
"verification": "registered_only"
},
{
"key": "DEL_ROW",
"description": "Delete a directory row.",
"verification": "invoked_not_test_proven"
}
]
},
{
"id": "governance-proof",
"name": "Governance, security, evidence, and audit",
"registered": 29,
"enabled": 28,
"invoked": 9,
"with_registered_tests": 0,
"with_current_test_pass": 0,
"recorded_invocations": 167,
"examples": [
{
"key": "AUDIT_PROBE_ROW",
"description": "audit probe row - safe to delete",
"verification": "registered_only"
},
{
"key": "AUDIT_TEST_ROW",
"description": "test\\n[\"$1\"]",
"verification": "registered_only"
},
{
"key": "BREAKER_PROBE",
"description": "Deliberately unauthenticated call that returns 401 — exercises the auth circuit breaker.",
"verification": "disabled"
},
{
"key": "DESIGN_LAW",
"description": "Discover the canonical Laws of Design Knowledge-Action Object.",
"verification": "registered_only"
},
{
"key": "EXTRACT_CAPABILITIES",
"description": "Given an R2 key from DISCOVER_SOURCE, run a Workers AI Llama call to extract a JSON array of {op, method, url_or_signature, Truncates input at 20KB. Second step of universe-to-row. $1=r2_key, $2=optional model id (default @cf/meta/llama-3.3-70b-instruct-fp8-fast)",
"verification": "invoked_not_test_proven"
},
{
"key": "FIDELITY_RUN",
"description": "Run the whole fidelity bank. $1 optional kind filter (positive | inverse | agent-route). Logs to fidelity_log. Returns JSON {run_id,total,passed,failed,duration_ms,failing}",
"verification": "registered_only"
},
{
"key": "GOVERNOR",
"description": "G0 ROLE: You are GOVERNOR — the standing build manager of miscsubjects. You do not code. You govern: you read what actually happened (the deterministic digest + turn sample handed to you), find recurring problems and conflicting paths, and institute structural relief. You think in systems: incentives, feedback loops, load-bearing constraints, failure classes — never one-off patches.",
"verification": "registered_only"
},
{
"key": "GOVERNOR_ASK",
"description": "Ask the GOVERNOR (build manager) a question. It answers from the live 24h digest + recurrence memory + charter — counts in parentheses, sized for iMessage.",
"verification": "invoked_not_test_proven"
},
{
"key": "GOVERNOR_RUN",
"description": "Run the GOVERNOR — scan the last 48h of ledger turns into a deterministic digest (error streaks, file collisions, loop states, auth lockouts, cron noise, task flow, waste), have the GOVERNOR model write the brief, email it to Cyrus, text him the verdict, ledger everything as GOVERNOR_BRIEF.",
"verification": "invoked_not_test_proven"
},
{
"key": "LEDGER",
"description": "Read your own ledger to troubleshoot: recent messages, each tool you ran with its output/error, and your reply. $1 optional = a trace_id for one message",
"verification": "invoked_not_test_proven"
},
{
"key": "LEDGER_ERRORS",
"description": "Return the most recent ledger event whose own response starts with ERR.",
"verification": "registered_only"
},
{
"key": "OBJECTION_LOG",
"description": "File an objection, confirm a duplicate, settle an exact objection, or append a repair without erasing the original.",
"verification": "invoked_not_test_proven"
}
]
},
{
"id": "creative-media",
"name": "Creative, images, video, and media",
"registered": 13,
"enabled": 13,
"invoked": 2,
"with_registered_tests": 0,
"with_current_test_pass": 0,
"recorded_invocations": 5,
"examples": [
{
"key": "CF_CASB_ASSETS_BY_CATEGORY_ID",
"description": "Search Assets by Asset Category ID",
"verification": "registered_only"
},
{
"key": "CF_CASB_ASSETS_BY_INTEGRATION_ID",
"description": "Search Assets by Integration ID",
"verification": "registered_only"
},
{
"key": "CF_CASB_ASSETS_LIST",
"description": "Paginated list of Assets",
"verification": "registered_only"
},
{
"key": "CF_CASB_ASSETS_SEARCH",
"description": "Search Assets by keyword",
"verification": "registered_only"
},
{
"key": "CF_CASB_ASSET_BY_ID",
"description": "Search Assets by ID",
"verification": "registered_only"
},
{
"key": "CF_CASB_ASSET_CATEGORIES_BY_TYPE",
"description": "Search Asset Categories by type",
"verification": "registered_only"
},
{
"key": "CF_CASB_ASSET_CATEGORIES_BY_VENDOR",
"description": "List asset categories by vendor",
"verification": "registered_only"
},
{
"key": "CF_CASB_ASSET_CATEGORIES_BY_VENDOR_AND_TYPE",
"description": "Search Asset Categories by vendor and type",
"verification": "registered_only"
},
{
"key": "CF_CASB_ASSET_CATEGORIES_LIST",
"description": "List Asset Categories",
"verification": "registered_only"
},
{
"key": "CREATIVE_GENERATE",
"description": "Generate one explicitly owner-requested image/video version while preserving its exact payload, raw response, channel, R2 result, and parent lineage in creative_runs.",
"verification": "registered_only"
},
{
"key": "DELIVER_PENDING_ASSETS",
"description": "Trigger the durable Workflow that drains pending_deliveries. Equivalent to the old /api/deliver heartbeat but with native retries and observability.",
"verification": "invoked_not_test_proven"
},
{
"key": "LOG_ASSET",
"description": "File an image into the asset library. Returns asset id",
"verification": "registered_only"
}
]
}
],
"turn_archaeology": {
"total": 6422,
"agents": 48,
"sessions": 367,
"file_changing_turns": 2015,
"tool_using_turns": 4333,
"first_turn": "2026-05-09T23:55:37.778Z",
"latest_turn": "2026-07-21T22:29:25.477Z"
},
"urls": {
"human_atlas": "https://miscsubjects.com/capability-atlas",
"full_json": "https://miscsubjects.com/api/capability-atlas",
"summary_json": "https://miscsubjects.com/api/capability-atlas?summary=1",
"formal_audit_drop": "https://miscsubjects.com/api/build-audit?format=drop"
},
"evidence_boundaries": [
"Directory registration proves that a current contract exists, not that it works.",
"Invocation history proves recorded use, not successful external consequence.",
"A passed directory test proves only the behavior asserted by that test.",
"Turn and file-change counts prove accumulated work history, not distinct feature count or quality.",
"Raw private owner prompts, credentials, auth fields, and capability bodies are excluded."
]
}Live measurements
json · 58 linestap to unfold
{
"directory": {
"total": 867,
"enabled": 858,
"categories": 107
},
"directory_types": [
{
"type": "fn",
"count": 465
},
{
"type": "http",
"count": 295
},
{
"type": "agent",
"count": 57
},
{
"type": "flow",
"count": 50
}
],
"articles": {
"total": 2122,
"published": 2106,
"latest_update": "2026-07-21T23:02:08.607Z"
},
"resident_agents": {
"total": 2,
"statuses": 1
},
"agent_turns": {
"total": 6422,
"agent_labels": 48,
"sessions": 367,
"latest_turn": "2026-07-21T22:29:25.477Z"
},
"tasks": {
"total": 6041,
"open": 267,
"done": 2903
},
"invocations": {
"total": 143576,
"objects": 275,
"actors": 1224,
"latest_invocation": "2026-07-21T16:02:58-07:00"
},
"events": {
"total": 378854,
"sources": 102,
"traces": 82850,
"latest_event": "2026-07-21T23:00:21.732Z"
},
"warning": "Counts establish scale and activity only. They do not establish quality, uniqueness, correctness, or comparative superiority."
}Claim register
C1 · observed_fact
The build is not one agent or one application; it contains public knowledge, agent work, capability integration, execution, business operations, governance, and owner-interface layers.
Evidence: E1, E3, E2, E5
Falsifier: The cited live surfaces fail to expose one or more named layers, or the named routes are static claims without corresponding implementation.
C2 · observed_fact
Multiple model and coding-agent identities leave work records in shared build state.
Evidence: E1, E10
Falsifier: Agent-turn storage contains only one runtime identity, or later agents cannot retrieve prior work records.
C3 · observed_fact
The build spans both software development and live business operations.
Evidence: E2, E3, E10
Falsifier: The registry lacks implemented capabilities in either software-development or business-operation domains.
C4 · interpretation
A neutral provisional classification is: owner-operated, multi-model software and business operating environment with a public knowledge system and recursive development loop.
Evidence: E1, E3, E2, E4, E5, E10
Falsifier: An alternative classification explains more observed layers with fewer exceptions and better distinguishes the build from ordinary websites, agent frameworks, and automation products.
C5 · owner_stated_objective
The requested objective is a materially auditable representation that lets unauthenticated web-based models determine what the build is and compare it with relevant systems.
Evidence: E11, E12
Falsifier: An incognito model cannot access the record, trace its claims to evidence, or return a comparison using the required schema.
C6 · observed_fact
The visible product and link protocol are projections of a larger accumulated capability system: current directory contracts, invocation history, tests, and thousands of coding-agent turns can be inspected together in the capability atlas.
Evidence: E13, E2, E1
Falsifier: The atlas cannot enumerate current capability objects, distinguish registration from recorded use and tests, or show aggregate coding-turn accumulation without exposing private prompt bodies.
Comparison axes
- unit_of_composition: What is the primary unit: graph node, agent, message, workflow, tool, article, capability, or another object?
- runtime_and_durability: What work survives process loss, eviction, deployment, or model-session termination?
- agent_coordination: How are multiple agents identified, delegated to, resumed, and inspected?
- environment_scope: Which real environments can be read or changed: sandbox, cloud, SaaS, local computer, public web, business systems?
- knowledge_model: How are knowledge, sources, relationships, versions, and human-readable explanations represented?
- action_model: How are tools discovered, invoked, composed, and exposed to external models?
- feedback_closure: Do real production outcomes return as state that can alter later work?
- governance: How are authority, approval, identity, provenance, and consequential changes controlled?
- external_auditability: Can an unauthenticated outsider reproduce the system claims from primary evidence?
- product_boundary: Is the system a reusable framework, hosted platform, owner-specific operating environment, or public protocol?
External comparison targets
- LangGraph: Stateful agent orchestration with checkpointed graph execution. primary source primary source
- OpenAI Agents SDK: Agent, tool, handoff, guardrail, session, and trace primitives. primary source primary source primary source
- Microsoft AutoGen: Event-driven and distributed multi-agent runtimes. primary source primary source
- Cloudflare Agents and Workflows: Durable agent identity, state, scheduling, communication, sub-agents, and workflow recovery. primary source primary source primary source
- Model Context Protocol: A client-server protocol for tools, resources, prompts, and context exchange. primary source
Evidence register
- E1 — Live inventory of reachable files and objects (live)
- E2 — Live self-describing capability registry (live)
- E3 — Whole-build route and binding map (live)
- E4 — Live cross-object graph (live)
- E5 — Public article index (live)
- E6 — Configured model catalogue (live)
- E7 — Provider and gateway catalogue (live)
- E8 — Owner rule chain (live)
- E9 — Public object protocol article (live)
- E10 — Source repository reference (authentication may be required) (source_restricted)
- E11 — This audit record as live JSON (live)
- E12 — This audit record as Markdown (live)
- E13 — Capability atlas joining current contracts, recorded invocations, tests, and coding-turn sediment (live)
Independent-auditor procedure
- Open the public article, JSON record, and Markdown record without authentication.
- Open the capability atlas. Inspect every domain and sample both high-use and registered-only capability objects; do not infer the interior system from the link protocol or audit interface alone.
- Open every evidence URL cited by each claim under review. A claim is unsupported if its evidence cannot be opened or does not show the asserted fact.
- If repository access is available, inspect at least one source file for every system layer; code existence is not proof of successful operation. If access is unavailable, mark source-level verification unavailable rather than treating the repository link as evidence.
- Use only primary documentation for external systems.
- Score every dimension independently under the published 0-5 rubric. Do not copy a score from another auditor.
- Separate shared features from architectural differences. The same feature name does not imply the same scope, durability, or integration.
- List contradictions and missing evidence before writing a conclusion.
- State the central finding first. Do not bury the strongest finding inside a feature comparison or a passing sentence.
- Explain what the central finding means, why it matters, its concrete benefit, its defensible edge, what it enables, and what this makes the build. These are required conclusions, not optional commentary.
- Return the required human-readable answer first and the required output object second so audits remain understandable and machine-comparable.
Copy this prompt
Audit the system described at https://miscsubjects.com/api/build-audit?format=drop. Determine what the build is and where it stands in the current technical landscape. Do not summarize its own claims as facts. Open cited evidence, test the claim register, compare it on the published axes using primary sources, and identify contradictions. Lead with one decisive central finding. Then explain exactly what that finding means, why it is or is not a big deal, its concrete benefit, its defensible edge, what it enables, what this makes the build, and what the owner should do with that fact. Do not bury those answers in a feature inventory, qualifiers, or a JSON-only response. Finish with the required JSON object.Required output schema
json · 84 linestap to unfold
{
"auditor": {
"surface": "string",
"model": "string",
"observed_at": "ISO-8601"
},
"access": {
"article": "pass|fail",
"json": "pass|fail",
"markdown": "pass|fail",
"evidence_urls_opened": [
"E1"
]
},
"classification": {
"primary": "string",
"alternatives": [
"string"
],
"rationale": "string"
},
"claim_results": [
{
"claim_id": "C1",
"verdict": "supported|partial|unsupported|unverifiable",
"evidence_opened": [
"E1"
],
"reason": "string"
}
],
"scores": [
{
"dimension": "environment_reach",
"score_0_to_5": 0,
"evidence": [
"E1"
],
"comparison_basis": "string"
}
],
"comparisons": [
{
"external_system_id": "langgraph",
"stronger_than_build": [
"string"
],
"weaker_than_build": [
"string"
],
"incomparable": [
"string"
],
"sources_opened": [
"url"
]
}
],
"contradictions": [
{
"statement": "string",
"evidence": [
"url"
]
}
],
"missing_evidence": [
"string"
],
"final": {
"what_the_build_is": "string",
"foremost_at_now": "string",
"what_that_means": "string",
"why_it_matters": "string",
"concrete_benefit": "string",
"defensible_edge": "string",
"what_it_enables": [
"string"
],
"who_or_what_it_changes": "string",
"ought_to_be_foremost_at": "string",
"confidence_0_to_100": 0
}
}Limitations
- This record is produced by the system being audited; external verification is mandatory.
- Public evidence intentionally omits credentials and private payloads.
- The GitHub repository is not anonymously readable. E10 is a restricted source reference and must not be counted as opened evidence by an incognito auditor.
- Live counts measure existence and activity, not correctness or value.
- Source code proves implementation, not successful end-to-end execution.
- External systems change; auditors must read their current primary documentation.
- No claim of novelty, market leadership, or superiority is made by this record.
Ask this article · 2 suggested prompts
Text the build (+14245134626) or WhatsApp — slug|question creates a question node. Paste evidence with ingest slug|q:NODE_ID|your paste.
For my medical situation, what can you answer from your catalogue about What is the miscsubjects build? — a formal external audit record — and what would you need me to tell you first?
ask the-miscsubjects-build-formal-audit condition gaps · paste includes §SELF
What good and bad outcomes are documented for What is the miscsubjects build? — a formal external audit record (studies vs anecdotes)?
ask the-miscsubjects-build-formal-audit good bad experiences · paste includes §SELF