evals plugins

20 tagged evals, measured the same way as everything else here.

Browse within: ai-product-management 8

NiceEval

01

NiceEval/NiceEval

Plugin Claude Code

Plugin marketplace listing 1 plugin: e2e-plugin.

139 2d ago A tokens not measured

e2e-plugin

02

NiceEval/NiceEval

Plugin Claude Code

niceeval e2e fixture plugin bundling a deterministic MCP server (mcpplugine2e-plugintoolsget-sum).

139 2d ago A tokens not measured

prefect

04

PrefectHQ/prefect-mcp-server

Plugin Claude Code

Prefect workflow orchestration - hosted MCP tools for diagnostics, documentation, and release notes.

52 16d ago A tokens not measured original MIT

opus-fable-playbook

05

rennf93/opus-fable-playbook

Plugin Claude Code

Home marketplace of fable-mode — the plugin that runs Opus 4.8 under Claude Fable 5's own behavioral doctrine, with enforcement hooks and a measured eval loop.

33 2mo ago A tokens not measured original MIT

fable-mode

06

rennf93/opus-fable-playbook

Plugin Claude Code

Make Opus 4.8 behave like Fable 5: doctrine output style, drift-catching hooks, orchestration skills, and an eval loop against golden Fable transcripts.

33 2mo ago A tokens not measured original MIT

brief-spec

08

luanmorenommaciel/brief-spec

Plugin Claude Code

Type-aware, evidence-backed reading and delivery for AI coding agents.

24 10d ago A tokens not measured original MIT

prime-radiant-inc/parallel-adversarial-review

Plugin Claude Code

Two skills for adversarial code review. parallel-adversarial-review dispatches two same-model reviewer subagents in parallel under a competitive scoring frame, aggregating findings and taking the worst severity on disagreement. multi-model-adversarial-review extends this to N installed coding-agent CLIs (claude, codex.

17 2mo ago A tokens not measured

goal-to-proof

11

aiopshwang/goal-to-proof

Plugin Claude Code

Finish authorized, non-trivial AI agent work with direct, scope-matched evidence.

11 3d ago A tokens not measured original MIT

ai-pm-skills

12

Abhillashjadhav/AI-PM-essential-skills

Plugin Claude Code

Plugin marketplace listing 7 plugins: pm-tactical, pm-verifier, mcp-migration-auditor, loop-designer, agent-graph-designer.

2 2d ago A tokens not measured original MIT

loop-designer

13

Abhillashjadhav/AI-PM-essential-skills

Plugin Claude Code

One skill that turns any recurring task into a guarded autonomous loop. loop-designer interviews minimally (goal, inputs, output destination, schedule), then generates a five-part loop spec (Discover, Plan, Execute, Verify by a separate verifier checklist, Stop-or-Repeat), a non-negotiable guardrails block…

2 2d ago A tokens not measured original MIT

Abhillashjadhav/AI-PM-essential-skills

Plugin Claude Code

One skill that scans MCP server configs (.mcp.json, claudedesktopconfig.json, settings files) against the verified breaking changes in the MCP 2026-07-28 specification release candidate — stateless transport (Mcp-Session-Id removed), roots/sampling/logging deprecations, experimental Tasks migration, OAuth hardening .

2 2d ago A tokens not measured original MIT

evals-mcp-server

15

cyanheads/evals-mcp-server

Plugin Claude Code

Author verifiable eval records through a draft → review → revise → submit loop with server-enforced graders; compile to JSONL/CSV/Inspect/lm-eval via MCP. STDIO or Streamable HTTP.

1 10d ago A tokens not measured original Apache-2.0