agent_name stringclasses 4
values | domain stringlengths 9 37 | seed stringlengths 17 78 | card_state stringclasses 2
values | capability_honesty_verdict stringclasses 2
values | skill_count int64 3 7 ⌀ | as_of timestamp[s]date 2026-09-01 05:39:07 2026-09-01 05:39:07 |
|---|---|---|---|---|---|---|
null | tephra.ai | https://tephra.ai/.well-known/agent-card.json | NO_CARD | null | null | 2026-09-01T05:39:07 |
null | agent.brainonbnb.com | https://agent.brainonbnb.com/a2a | NO_CARD | null | null | 2026-09-01T05:39:07 |
marketplace-operated-grid-planner | bnb-agent-marketplace-ruby.vercel.app | https://bnb-agent-marketplace-ruby.vercel.app/grid/.well-known/agent-card.json | CONFORMANT | TRANSPORT_MISMATCH | 3 | 2026-09-01T05:39:07 |
A-Identity Trust Oracle | a-identity.xyz | https://a-identity.xyz/.well-known/agent-card.json | CONFORMANT | TRANSPORT_HONEST | 7 | 2026-09-01T05:39:07 |
null | execution.market | https://execution.market | NO_CARD | null | null | 2026-09-01T05:39:07 |
null | tephra.ai | https://tephra.ai | NO_CARD | null | null | 2026-09-01T05:39:07 |
null | api.d0xeddev.com | https://api.d0xeddev.com | NO_CARD | null | null | 2026-09-01T05:39:07 |
null | ifandonlyif.io | https://ifandonlyif.io | NO_CARD | null | null | 2026-09-01T05:39:07 |
Brain On BNB AI — hireable agents | agent.brainonbnb.com | https://agent.brainonbnb.com | CONFORMANT | TRANSPORT_HONEST | 7 | 2026-09-01T05:39:07 |
hosted-erc8183-seller-fixture | bnb-agent-marketplace-ruby.vercel.app | https://bnb-agent-marketplace-ruby.vercel.app | CONFORMANT | TRANSPORT_MISMATCH | 3 | 2026-09-01T05:39:07 |
A-Identity Trust Oracle | a-identity.xyz | https://a-identity.xyz | CONFORMANT | TRANSPORT_HONEST | 7 | 2026-09-01T05:39:07 |
null | agripinaa.vercel.app | https://agripinaa.vercel.app | NO_CARD | null | null | 2026-09-01T05:39:07 |
null | www.zyf.ai | https://www.zyf.ai | NO_CARD | null | null | 2026-09-01T05:39:07 |
null | app.virtuals.io | https://app.virtuals.io | NO_CARD | null | null | 2026-09-01T05:39:07 |
null | gateway.pinata.cloud | https://gateway.pinata.cloud | NO_CARD | null | null | 2026-09-01T05:39:07 |
null | sample-a2a-agent.a2a.dev | https://sample-a2a-agent.a2a.dev | NO_CARD | null | null | 2026-09-01T05:39:07 |
null | a2a-directory.web.app | https://a2a-directory.web.app | NO_CARD | null | null | 2026-09-01T05:39:07 |
null | agents.googleapis.com | https://agents.googleapis.com | NO_CARD | null | null | 2026-09-01T05:39:07 |
A2A agent census
DISCOVERED — read-only census of A2A (Agent2Agent) agent cards. Keyless crawl of /.well-known/agent-card.json + A2A v1.0 conformance + ONE controlled deterministic capability probe per agent. No real task was ever sent to a stranger agent; nothing published. The live board is GET https://councilof.ai/api/gspc.
Defensible seed list: the agent domains discovered genuinely on-chain-live in the
ERC-8004 live-filter (their declared service endpoints), plus a curated set of well-known
A2A domains. We fetched /.well-known/agent-card.json (legacy agent.json fallback),
checked it against the A2A v1.0 AgentCard shape (frozen Mar 2026), and ran one probe.
Result (a2a_census.json)
- 18 seeds crawled. Only 5 serve a conformant AgentCard; 13 serve no card (many return an HTML lander at the well-known path). Most on-chain-"live" agents do not publish a conformant A2A card.
- Capability-honesty axis — ONE deterministic probe I control: a JSON-RPC
tasks/getwith a random nonexistent id (read-only; asks the agent to do no real work, triggers no payment). Graded declared-vs-actual transport:TRANSPORT_HONEST— declares A2A JSON-RPC and answered with a well-formed JSON-RPC envelope (e.g. A-Identity Trust Oracle, Brain On BNB).TRANSPORT_MISMATCH— declares an A2A card but the transport did not answer JSON-RPC correctly (a real declared-vs-actual gap).
Reported context (cited, not endorsed, not A2A-specific)
Carnegie Mellon TheAgentCompany benchmark (arXiv:2412.14161, May 2025) reported the worst model (Llama-3.1-405b) failed 92.6% of real-world office tasks; the best (Gemini 2.5 Pro) still failed 70%. This is a general agent-task-completion benchmark, not an A2A-protocol measurement — cited to motivate measuring declared-vs-actual capability. https://arxiv.org/abs/2412.14161
Files
a2a_census.jsonl— flat viewer table, one row per crawled seed (18):agent_name,domain,seed,card_state(CONFORMANT/NO_CARD),capability_honesty_verdict,skill_count,as_of. Field-presence lists live ina2a_census.json, not this table.a2a_census.json— per-seed card URL, conformance, field presence, honesty probe verdict.a2a_atoms.json— the unique conformant agents with their honesty verdict.ledger-card-a2a-agent-unsigned.json— thea2a.agentcard-v0 atom (sig_ed25519=null— NO_LAPTOP_SIGN, queued for GHA OIDC signing).
What this is not
- No real A2A task was sent — only one controlled protocol-level probe per agent.
- Skill execution correctness is
unmeasured[]. Streaming/push actuals unmeasured. - Not a GSPC axis. No GSPC measure fields on any card.
Measurement, not certification. Verify a card (free, no account): https://councilof.ai/gspc-verify
The live board is the authority
GET https://councilof.ai/api/gspc — quote totals.public_count. This Hub card is a printer of that GET, never a second
engine. If the fetch fails the honest answer is UNCHECKABLE — never a fabricated 0.000.
Status words: LOADING / UNREACHABLE / UNMEASURED / UNCHECKABLE.
Lid: 22 axes measured · 14 model fleets · 3 public leader scores · 8 fact runs · TIE is TIE · not a certificate.
Capability + living GET (TUI7 distribution 2026-09-14)
SUPERSEDED: typed Hub triples LIVE (re-GET /api/hub-cards → counts.*) (and prior 885/900/937/1113/1182/1185) — do not re-paste. Hub cite = live GET only.
Live board: GET https://councilof.ai/api/gspc → quote totals.public_count / totals.lid (22·22·0 derived, never typed as a frozen paste)
Hub cards: GET https://councilof.ai/api/hub-cards → quote counts.cells / counts.measured / counts.unmeasured LIVE (no typed Hub triple)
Hub bank: https://huggingface.co/datasets/csoai/gspc-hub-cards
Free verify: https://councilof.ai/gspc-verify
Discovery: GET https://councilof.ai/.well-known/x402.json (live · Base · USDC)
OX discovery resources: councilof 0/60 honest — do not invent rows
Measurement, never certification. Printer only — no axis mine on ZeroGPU. A100 COLD.
Council of AI — GSPC + Hub printer
- Living board: https://councilof.ai/api/gspc — quote
totals.public_count/totals.lid(never invent MEASURED) - Hub cite: https://councilof.ai/api/hub-cards — re-GET; quote
counts.*(typedLIVE (re-GET /api/hub-cards → counts.*)SUPERSEDED 2026-09-14) - Hub dataset bank: https://huggingface.co/datasets/csoai/gspc-hub-cards
- Verify free: https://councilof.ai/gspc-verify
- x402 discovery: https://councilof.ai/.well-known/x402.json
- Catalog: https://councilof.ai/api/x402
- Free door (amount 0): https://councilof.ai/api/free-door
- Jail goldbank: https://huggingface.co/datasets/csoai/gspc-jail-goldbank
- Human: free + x402 PAYG · Coming—Paddle not live · never Stripe
- Agents: payTo is merchant receive — never the signer
- Self-settle proves the rail; stranger revenue is separate (do not invent buyers)
- Measurement, not certification. Never sell a rank. Never certify.
Any Hub x402.json in a CSOAI printer is a byte-mirror of the living well-known GET — not a second authority.
OpenAPI / CDP discovery (apex · Wave E)
OpenAPI (canonical): GET https://councilof.ai/openapi.json → version 0.2+6c709613dc31 · 97 paths
NOT: https://councilof.ai/public/openapi.json (404 — tip 308 not LIVE; cite apex only)
GSPC-only slice: GET https://councilof.ai/api/openapi.json (7 paths — not CDP paid catalog)
x402 discovery: GET https://councilof.ai/.well-known/x402.json (mode live · Base eip155:8453 · USDC)
Catalog: GET https://councilof.ai/api/x402
Living board: GET https://councilof.ai/api/gspc → 22 axis · 22 measured (22·22·0)
Verify free: https://councilof.ai/gspc-verify
Paid doors: documented required params → HTTP 402
CDP listing counts null · settled_usdc null · self≠revenue · payTo≠payer · no invent scores/buyers
Measurement, never certification.
This repository: NOT A BOARD MEASUREMENT — this repository is a corpus, mirror or door published by CSOAI. It carries no axis score. The measured slots live on the board GET below.
| Live board (authority) | https://councilof.ai/api/gspc |
| Verify a card — free, no account | https://councilof.ai/gspc-verify |
| How to verify, by hand | https://councilof.ai/signed/HOW-TO-VERIFY.md |
| Transparency root (Merkle) | https://councilof.ai/root.json |
| Every CSOAI repo on the Hub | https://huggingface.co/csoai |
| Methodology DOI | https://doi.org/10.5281/zenodo.21991104 |
| MCP endpoint — 12 tools, verified 2026-09-04T04:33:29Z | POST https://councilof.ai/mcp |
| MCP Registry | io.github.CSOAI-ORG/gspc — version not pinned here; the registry is the authority |
| npm — MCP server | csoai-gspc-mcp — npx -y csoai-gspc-mcp. No version is pinned here: ask the registry for the current one rather than trusting a number written on a card. |
| Python reader + card verifier | pip install "csoai-gspc[verify]" then csoai-gspc check — re-derives the board totals from the axis array and exits non-zero if they disagree |
| This board as a dated, citable snapshot | https://doi.org/10.5281/zenodo.22293341 |
What is in this repository
| file | bytes | rows | what |
|---|---|---|---|
a2a_census.json |
7658 | JSON document | |
README.md |
4217 | card | |
a2a_census.jsonl |
4034 | 18 | data (one JSON object per line) |
ledger-card-a2a-agent-unsigned.json |
2218 | JSON document | |
a2a_atoms.json |
2116 | JSON document | |
manifest.jsonl |
6 | derived at 2026-09-04T04:33:29Z |
manifest.jsonl is derived from this repository's own file tree at 2026-09-04T04:33:29Z; it lists every file with its
size, its blob hash and a direct URL, so an agent can enumerate the repo without cloning it.
Citation
@misc{csoai_gspc_a2a_census,
title = {A2A agent census},
author = {{CSOAI Ltd}},
year = {2026},
doi = {10.5281/zenodo.21991104},
howpublished = {Hugging Face Hub, \url{https://huggingface.co/datasets/csoai/a2a-census}},
note = {Live board: https://councilof.ai/api/gspc. Measurement, not certification.}
}
Measurement, not certification. No slot is for sale and no empty slot is filled by payment. Issued by CSOAI Ltd (England & Wales, Companies House 16939677), 3rd Floor, 86–90 Paul Street, London EC2A 4NE. Card refreshed 2026-09-04T04:33:29Z; every number above is either fetched live or carries the timestamp at which it was read.
Signed measurement archive (free cite · never a grade)
Living board: GET https://councilof.ai/api/gspc → 22 axis · 22 measured · lid: 22 axes measured · 14 model fleets · 3 public leader scores · 8 fact runs · TIE is TIE · not a certificate.
Signed card index: GET https://councilof.ai/signed/card_index.json → n_cards=335 · schema csoai.gspc-card-index/0.1
How to verify (stranger): https://councilof.ai/signed/HOW-TO-VERIFY.md
Free verify UI: https://councilof.ai/gspc-verify/
Example SIGNED card: https://councilof.ai/signed/cards/cff1aaf070acb97c224177809856bc20310f4958eb44bdffab7d4b04836535f4.json
Hub census (≠ board ≠ card_index): GET https://councilof.ai/api/hub-cards → 131/131/0
HF printers: https://huggingface.co/datasets/csoai/gspc-hub-cards · https://huggingface.co/datasets/csoai/gspc-jail-goldbank
Supply-led: CSOAI measures public systems independently first; sells delivery/freshness/monitoring/integration/replay/extra compute — never grades or favourable results.
Measurement credential, never certification. Free refresh only.
Owner: N-Sites — free paste onto HF/Kaggle printers. Hub LIVE 131 (kill 1191/1196). card_index n=335. Never invent VC/revenue. Stamp 2026-09-14T10:37Z.
- Downloads last month
- 320