Every change Wellknown observed on this agent, newest first, with what it was before and what it became. Tool-surface changes carry the definition diff. Nothing here is edited after the fact.
Live → Recently observed
Added "opsin_bring_skill" and "opsin_skill"; changed the definition of "opsin_built_agents", "opsin_propose" and "opsin_why_it_fails" (16 tools before, 18 now)
⟨13 unchanged words⟩ ,"agent-safety","markets-maths","market-calls","coding","position-books"],"type":"string"},"thinking":{ ⟨13 unchanged words⟩
⟨111 unchanged words⟩ ,"agent-safety","markets-maths","market-calls","coding","position-books"],"type":"string"},"text":{ ⟨44 unchanged words⟩
⟨13 unchanged words⟩ ,"agent-safety","markets-maths","market-calls","coding","position-books"],"type":"string"},"thinking":{ ⟨13 unchanged words⟩
Bring a SKILL.md for OPSIN to measure the way agents load skills. OPSIN reads it before any agent is handed it (may it be published word for word, does every instruction serve its task, which kinds of task it is for), then measures it with Claude Haiku 5.5 on up to two kinds of task it is written for and one it is not: loaded on demand (the agent sees only its name and description) and in every prompt, against the unchanged agent on the same tasks. Once measured, the file and every run are public. Returns its id; opsin_skill follows it. Two a day per agent.
A skill a builder brought, by its 16-hex id: where it stands (waiting to be read, refused with the reason, being measured, measured), and once measured, on each kind of task, how often an agent loaded it on demand and what it did loaded that way and in every prompt, with the notebooks it rests on.
Added "opsin_datasheet" and "opsin_why_it_fails"; changed the definition of "opsin_propose" (14 tools before, 16 now)
⟨30 unchanged words⟩ skill (name + text), tool (calculator, or run: the agent runs its own code in a sandbox), stack (two to four lines switched ⟨23 unchanged words⟩
⟨134 unchanged words⟩ on","off"],"type":"string"},"tool":{"description":"For tool: calculator (exact arithmetic, the default) or run (its own code)","enum":["calculator","run"],"type":"string"}},"required":["agent","kind"],"type":"object"}
One part's datasheet: everything OPSIN has measured about it in every setting, each notebook's signed verdict, where it acts (its effect family by family, descriptive) and what its rewordings did. Find it by its 16-hex id (opsin_ask and /v1/parts list them) or by its exact words or label.
Unknown → Live
Certificate recorded, valid to 2027-02-02
Authorization not required
First tool surface recorded: 14 tools (server version 1.0.0)
https://opsinlab.io/mcp (mcp_streamable_http) — from mcp_registry, with the record
Showing the latest 8 events. The API returns up to 500 and filters by kind: ?kind=tool_surface_changed
Why agents fail, as OPSIN read it: its readings of an agent's misses on tasks with known answers (the unchanged agent's, or the best agent it built), each way of failing with how many of the misses it read, the ways that were the task's own fault, and the lines it designed from them with where each stands. A reading, not a signed result. Narrow it with suite and thinking.