studio-insight
claudeBusiness analysis toolkit — personas, journey maps, process flows, domain modeling
6 skills · 6 commands · 12 agents
3
claude
studio-evolve
Self-evolving ontology pipeline — ground the IR in data evidence, derive simulated cases, attribute failed trajectories to five-element deficits as atomic proposals, and admit them only through a dual-gate sandbox scored on three growth axes (taller / sturdier / wider)
/
Plugin studio-evolve from marketplace astra-studio in Ameng2001/astra-studio-plugins. Commands read from the repo's manifest — 6 days ago.
What it installs
Skills (4)
evolve-attributeMine one window of runtime trajectories into atomic ontology edit proposals — label each trajectory from user feedback, split correct vs wrong, distil commonalities from correct runs, locate the divergence point of wrong runs and attribute it to one of seven five-element deficit types, then write a
evolve-groundGround a five-element ontology IR in real data — for every Object / Mapping / function-backed Rule, issue a probe query against the declared data source, keep the sample as an Evidence row (IR §8), and mark unverifiable entries TODO(FDE). The dynamic-ontology-construction step of the self-evolving p
evolve-simulateDerive labelled competency cases from a grounded ontology — one or more cases per Rule (rejected), Action (committed), Link (multi-hop query) and Function (derived value), each with a difficulty tier, a scenario family, a private label and perturbation variants. Writes them to the sim case bank; sim
evolve-verifyAdjudicate one ontology edit proposal in a sandbox under the dual gate — copy the compiled plugin, apply the proposal, run source cases (inner gate: strictly better on three growth axes within a cost band) and a freshly resampled real-only held-out set (outer gate: no regression on any axis), then A
Slash commands (5)
/studio-evolve:attributeMine one window of runtime trajectories into atomic ontology edit proposals — divergence point → deficit type → single-hypothesis proposal. Does not adjudicate.
/studio-evolve:cycleRun one self-evolution round on an ontology — attribute trajectories into proposals, let the user review, verify each proposal under the dual gate, and replay accepted ones onto the live IR after confirmation.
/studio-evolve:groundGround every Object / Mapping / function-backed Rule of an ontology IR in real data via probe queries — appends IR §8 Evidence.
/studio-evolve:simulateDerive labelled, tiered competency cases from a grounded ontology (rules → rejected, actions → committed, links → multi-hop, functions → derived values) into the sim case bank. Sim cases feed the inner gate only.
/studio-evolve:verifyAdjudicate one proposal in a sandbox under the dual gate (inner: strictly better on source cases; outer: no regression on a resampled real-only held-out set) and record the three growth readings.
Subagents (1)
ontology-evolverEvolution analyst who turns agent failures into ontology hypotheses. Invoke during evolve-attribute to check that a divergence point really is the first step inconsistent with ground truth, that the deficit is attributed to the ontology (not the agent or the model), that the proposal carries a singl
More from Ameng2001/astra-studio-plugins
All 10 plugins →fund-review
claudeGovernment IT procurement quote optimization + financial review pre-audit — parse budget standard and quotes, produce standards-cited optimization suggestions, rewrite spreadsheets per approved plan, score review reports
12 skills · 4 commands · 2 agents
3
studio-docs
claudeDocument delivery suite — blueprints, writing experts, parallel chapter generation, multi-format export (MD/DOCX/PDF)
6 skills · 4 commands · 1 agent
3
studio-core
claudeStudio workspace management — init, promote, status
4 skills · 4 commands · hooks
3
studio-ontology
claudeFDE pipeline — compile business analysis (DDD/event-storm) into a runnable clife-onto-engine ontology plugin (five-element schema: Object/Link/Function/Rule/Action)
3 skills · 4 commands · 2 agents
3
studio-planner
claudePlugin planning pipeline — event storming, domain modeling, skill design, spec generation
5 skills · 1 command · 3 agents
3
Similar plugins
ecc
claudecodexHarness-native ECC operator layer - 71 agents, 314 skills, 98 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
200 skills · 98 commands · 71 agents · 1 MCP server · hooks
affaan-m/ECC 302,088
ruflo-metaharness
claudeMetaHarness integration — surfaces score/genome/mint/mcp-scan/threat-model via skills; pairs with @metaharness/router (ADR-148/149) for cost-optimal model routing; honors ADR-150 optional-augmentation constraint
13 skills · 1 command · 1 agent · hooks
ruvnet/ruflo 81,589
cli-anything
claudeBuild powerful, stateful CLI interfaces for any GUI application using the cli-anything harness methodology.
5 commands
HKUDS/CLI-Anything 58,665
adobe-for-creativity
claudeHarness Adobe's creative AI-powered tools to edit images, automate design workflows, and bring creative visions to life — from background removal to vectorization and professional retouching.
26 skills · 1 MCP server
anthropics/claude-plugins-official 47,076
llm-finetuning
claudecodexEval-gated LLM fine-tuning lifecycle (LoRA/QLoRA, DPO, GRPO/RLVR, vision SFT, quantized export). Use when fine-tuning an open-weights model end-to-end: building the eval harness first, selecting method and model, preparing data, training with Unsloth or TRL, gating checkpoints, and exporting GGUF/FP8.
10 skills · 2 commands · 3 agents
wshobson/agents 42,461
agent-harness
claudeTurn any domain folder of skills into a bounded agentic loop: a manifest builder inventories a domain's skills/tools/checks, a goal compiler turns a goal into a verifiable task plan (refusing vague goals with forcing questions), and a JSON-backed loop controller drives execute->verify->close with retry caps, controller-run verification (no verification theater), human escalation on exhausted budgets, and a close gate that refuses while any task is unverified or unwaived. Ships 3 stdlib Python tools, 18 committed per-domain harness manifests + JSON schema, 3 references citing the 2024-2026 agent-harness canon (Anthropic long-running harnesses, verifier's law, SWE-agent, Ralph loop, Cognition), harness-runner agent + /cs-harness command. Use when an agent or subagent should pick up a goal, define its tasks, complete and verify them, and close the loop.
1 skill · 1 command · 1 agent
alirezarezvani/claude-skills 29,725
Maintain studio-evolve?
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.
claude & codex plugins10 plugins