@evolvingmachines/evolve 0.0.66 → 0.0.67-train-harness-wave-2026-09-25.20260926.e00a8a6d

package/harness-capabilities.json

@@ -18,6 +18,40 @@
   ],
   "defaultReasoningEffort": "medium",
   "harnesses": {
+    "antigravity": {
+      "defaultModel": "gemini-3.8-flash",
+      "models": [
+        {
+          "alias": "gemini-3.8-flash",
+          "modelId": "vertex_ai/gemini-3.8-flash",
+          "description": "Latest Flash on Vertex AI: coding + agentic planning"
+        },
+        {
+          "alias": "gemini-3.5-flash-lite",
+          "modelId": "vertex_ai/gemini-3.5-flash-lite",
+          "description": "Latest Flash-Lite on Vertex AI: most cost-effective"
+        },
+        {
+          "alias": "gemini-3.1-pro-preview",
+          "modelId": "vertex_ai/gemini-3.1-pro-preview",
+          "description": "Latest Pro on Vertex AI: complex agentic + coding"
+        }
+      ],
+      "effortSupport": "level",
+      "efforts": [
+        "off",
+        "minimal",
+        "low",
+        "medium",
+        "high",
+        "xhigh",
+        "max",
+        "thinking"
+      ],
+      "defaultEffort": "high",
+      "supportsConfig": false,
+      "presets": []
+    },
     "claude": {
       "defaultModel": "opus",
       "models": [
@@ -236,6 +270,40 @@
       "supportsConfig": false,
       "presets": []
     },
+    "dsh": {
+      "defaultModel": "openrouter/deepseek/deepseek-v4.1-flash",
+      "models": [
+        {
+          "alias": "openrouter/deepseek/deepseek-v4.1-flash",
+          "modelId": "openrouter/deepseek/deepseek-v4.1-flash",
+          "description": "DeepSeek V4.1 Flash via OpenRouter"
+        },
+        {
+          "alias": "fireworks/deepseek-v4.1-flash",
+          "modelId": "fireworks/deepseek-v4.1-flash",
+          "description": "DeepSeek V4.1 Flash via Fireworks"
+        },
+        {
+          "alias": "openrouter/deepseek/deepseek-v4-pro-0813",
+          "modelId": "openrouter/deepseek/deepseek-v4-pro-0813",
+          "description": "DeepSeek V4 Pro via OpenRouter"
+        },
+        {
+          "alias": "fireworks/deepseek-v4-pro-0813",
+          "modelId": "fireworks/deepseek-v4-pro-0813",
+          "description": "DeepSeek V4 Pro via Fireworks"
+        }
+      ],
+      "effortSupport": "level",
+      "efforts": [
+        "low",
+        "medium",
+        "high"
+      ],
+      "defaultEffort": "high",
+      "supportsConfig": false,
+      "presets": []
+    },
     "gemini": {
       "defaultModel": "gemini-3.5-flash",
       "models": [
@@ -396,6 +464,184 @@
       "supportsConfig": false,
       "presets": []
     },
+    "pi": {
+      "defaultModel": "openrouter/anthropic/claude-opus-5",
+      "models": [
+        {
+          "alias": "openrouter/anthropic/claude-fable-5.1",
+          "modelId": "openrouter/anthropic/claude-fable-5.1",
+          "description": "Anthropic Fable 5.1 via OpenRouter"
+        },
+        {
+          "alias": "openrouter/anthropic/claude-opus-5",
+          "modelId": "openrouter/anthropic/claude-opus-5",
+          "description": "Anthropic Opus 5 via OpenRouter"
+        },
+        {
+          "alias": "openrouter/anthropic/claude-sonnet-5",
+          "modelId": "openrouter/anthropic/claude-sonnet-5",
+          "description": "Anthropic Sonnet 5 via OpenRouter"
+        },
+        {
+          "alias": "openrouter/anthropic/claude-haiku-4.5",
+          "modelId": "openrouter/anthropic/claude-haiku-4.5",
+          "description": "Anthropic Haiku via OpenRouter"
+        },
+        {
+          "alias": "openrouter/openai/gpt-6-astra",
+          "modelId": "openrouter/openai/gpt-6-astra",
+          "description": "OpenAI GPT-6 Astra via OpenRouter"
+        },
+        {
+          "alias": "openrouter/openai/gpt-5.6-sol",
+          "modelId": "openrouter/openai/gpt-5.6-sol",
+          "description": "OpenAI GPT-5.6 Sol via OpenRouter"
+        },
+        {
+          "alias": "openrouter/openai/gpt-5.6-terra",
+          "modelId": "openrouter/openai/gpt-5.6-terra",
+          "description": "OpenAI GPT-5.6 Terra via OpenRouter"
+        },
+        {
+          "alias": "openrouter/openai/gpt-5.6-luna",
+          "modelId": "openrouter/openai/gpt-5.6-luna",
+          "description": "OpenAI GPT-5.6 Luna via OpenRouter"
+        },
+        {
+          "alias": "openrouter/google/gemini-3.6-flash",
+          "modelId": "openrouter/google/gemini-3.6-flash",
+          "description": "Gemini 3.6 Flash via OpenRouter"
+        },
+        {
+          "alias": "openrouter/qwen/qwen3.7-max",
+          "modelId": "openrouter/qwen/qwen3.7-max",
+          "description": "Qwen 3.7 Max via OpenRouter"
+        },
+        {
+          "alias": "openrouter/moonshotai/kimi-k3",
+          "modelId": "openrouter/moonshotai/kimi-k3",
+          "description": "Kimi K3 via OpenRouter"
+        },
+        {
+          "alias": "openrouter/z-ai/glm-5.3",
+          "modelId": "openrouter/z-ai/glm-5.3",
+          "description": "Zhipu GLM-5.3 via OpenRouter"
+        },
+        {
+          "alias": "openrouter/z-ai/glm-5.3-flash",
+          "modelId": "openrouter/z-ai/glm-5.3-flash",
+          "description": "Zhipu GLM-5.3 Flash (OpenRouter id; the Evolve gateway serves it from Fireworks)"
+        },
+        {
+          "alias": "openrouter/deepseek/deepseek-v4.1-flash",
+          "modelId": "openrouter/deepseek/deepseek-v4.1-flash",
+          "description": "DeepSeek V4.1 Flash via OpenRouter"
+        }
+      ],
+      "effortSupport": "level",
+      "efforts": [
+        "off",
+        "minimal",
+        "low",
+        "medium",
+        "high",
+        "xhigh",
+        "max",
+        "thinking"
+      ],
+      "defaultEffort": "high",
+      "supportsConfig": false,
+      "presets": []
+    },
+    "prime-agent": {
+      "defaultModel": "openrouter/anthropic/claude-opus-5",
+      "models": [
+        {
+          "alias": "openrouter/anthropic/claude-fable-5.1",
+          "modelId": "openrouter/anthropic/claude-fable-5.1",
+          "description": "Anthropic Fable 5.1 via OpenRouter"
+        },
+        {
+          "alias": "openrouter/anthropic/claude-opus-5",
+          "modelId": "openrouter/anthropic/claude-opus-5",
+          "description": "Anthropic Opus 5 via OpenRouter"
+        },
+        {
+          "alias": "openrouter/anthropic/claude-sonnet-5",
+          "modelId": "openrouter/anthropic/claude-sonnet-5",
+          "description": "Anthropic Sonnet 5 via OpenRouter"
+        },
+        {
+          "alias": "openrouter/anthropic/claude-haiku-4.5",
+          "modelId": "openrouter/anthropic/claude-haiku-4.5",
+          "description": "Anthropic Haiku via OpenRouter"
+        },
+        {
+          "alias": "openrouter/openai/gpt-6-astra",
+          "modelId": "openrouter/openai/gpt-6-astra",
+          "description": "OpenAI GPT-6 Astra via OpenRouter"
+        },
+        {
+          "alias": "openrouter/openai/gpt-5.6-sol",
+          "modelId": "openrouter/openai/gpt-5.6-sol",
+          "description": "OpenAI GPT-5.6 Sol via OpenRouter"
+        },
+        {
+          "alias": "openrouter/openai/gpt-5.6-terra",
+          "modelId": "openrouter/openai/gpt-5.6-terra",
+          "description": "OpenAI GPT-5.6 Terra via OpenRouter"
+        },
+        {
+          "alias": "openrouter/openai/gpt-5.6-luna",
+          "modelId": "openrouter/openai/gpt-5.6-luna",
+          "description": "OpenAI GPT-5.6 Luna via OpenRouter"
+        },
+        {
+          "alias": "openrouter/google/gemini-3.6-flash",
+          "modelId": "openrouter/google/gemini-3.6-flash",
+          "description": "Gemini 3.6 Flash via OpenRouter"
+        },
+        {
+          "alias": "openrouter/qwen/qwen3.7-max",
+          "modelId": "openrouter/qwen/qwen3.7-max",
+          "description": "Qwen 3.7 Max via OpenRouter"
+        },
+        {
+          "alias": "openrouter/moonshotai/kimi-k3",
+          "modelId": "openrouter/moonshotai/kimi-k3",
+          "description": "Kimi K3 via OpenRouter"
+        },
+        {
+          "alias": "openrouter/z-ai/glm-5.3",
+          "modelId": "openrouter/z-ai/glm-5.3",
+          "description": "Zhipu GLM-5.3 via OpenRouter"
+        },
+        {
+          "alias": "openrouter/z-ai/glm-5.3-flash",
+          "modelId": "openrouter/z-ai/glm-5.3-flash",
+          "description": "Zhipu GLM-5.3 Flash (OpenRouter id; the Evolve gateway serves it from Fireworks)"
+        },
+        {
+          "alias": "openrouter/deepseek/deepseek-v4.1-flash",
+          "modelId": "openrouter/deepseek/deepseek-v4.1-flash",
+          "description": "DeepSeek V4.1 Flash via OpenRouter"
+        }
+      ],
+      "effortSupport": "level",
+      "efforts": [
+        "off",
+        "minimal",
+        "low",
+        "medium",
+        "high",
+        "xhigh",
+        "max",
+        "thinking"
+      ],
+      "defaultEffort": "high",
+      "supportsConfig": false,
+      "presets": []
+    },
     "qwen": {
       "defaultModel": "qwen3.7-max",
       "models": [
@@ -425,6 +671,45 @@
       "defaultEffort": "thinking",
       "supportsConfig": false,
       "presets": []
+    },
+    "zcode": {
+      "defaultModel": "openrouter/z-ai/glm-5.3",
+      "models": [
+        {
+          "alias": "openrouter/z-ai/glm-5.3",
+          "modelId": "openrouter/z-ai/glm-5.3",
+          "description": "Zhipu GLM-5.3 via OpenRouter"
+        },
+        {
+          "alias": "openrouter/z-ai/glm-5.3-flash",
+          "modelId": "openrouter/z-ai/glm-5.3-flash",
+          "description": "Zhipu GLM-5.3 Flash via OpenRouter"
+        },
+        {
+          "alias": "fireworks/glm-5.3",
+          "modelId": "fireworks/glm-5.3",
+          "description": "Zhipu GLM-5.3 via Fireworks"
+        },
+        {
+          "alias": "fireworks/glm-5.3-flash",
+          "modelId": "fireworks/glm-5.3-flash",
+          "description": "Zhipu GLM-5.3 Flash via Fireworks"
+        }
+      ],
+      "effortSupport": "level",
+      "efforts": [
+        "off",
+        "minimal",
+        "low",
+        "medium",
+        "high",
+        "xhigh",
+        "max",
+        "thinking"
+      ],
+      "defaultEffort": "high",
+      "supportsConfig": false,
+      "presets": []
     }
   }
 }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@evolvingmachines/evolve",
-  "version": "0.0.66",
+  "version": "0.0.67-train-harness-wave-2026-09-25.20260926.e00a8a6d",
   "keywords": [
     "ai",
     "agents",
@@ -49,14 +49,15 @@
     "build": "tsx scripts/generate-harness-capabilities.ts --check && tsup --minify && node scripts/copy-spec.mjs && node scripts/copy-skills.mjs",
     "generate:capabilities": "tsx scripts/generate-harness-capabilities.ts",
     "dev": "tsup src/index.ts --watch",
-    "type-check": "tsc --noEmit && tsc -p tsconfig.types-test.json",
+    "type-check": "tsc --noEmit && tsc -p tsconfig.types-test.json && node scripts/check-test-syntax.mjs",
     "test": "npm run test:unit && npm run test:integration",
     "pretest:unit": "npm run build",
-    "test:unit": "tsx tests/unit/semaphore.test.ts && tsx tests/unit/swarm-concurrency.test.ts && tsx tests/unit/swarm-verify.test.ts && tsx tests/unit/swarm-retry-verify.test.ts && tsx tests/unit/prompt-construction.test.ts && tsx tests/unit/observability-metadata.test.ts && tsx tests/unit/observability-identity.test.ts && tsx tests/unit/skills-integration.test.ts && tsx tests/unit/skills-resolver.test.ts && tsx tests/unit/skills-tag-peel.test.ts && tsx tests/unit/auth-config.test.ts && tsx tests/unit/config-validation.test.ts && tsx tests/unit/agent-presets.test.ts && tsx tests/unit/run-resume-option.test.ts && tsx tests/unit/native-config-delivery.test.ts && tsx tests/unit/browser-config.test.ts && tsx tests/unit/integrations-config.test.ts && tsx tests/unit/managed-secrets.test.ts && tsx tests/unit/managed-modal.test.ts && tsx tests/unit/plugins-config.test.ts && tsx tests/unit/session-runtime.test.ts && tsx tests/unit/cost-api.test.ts && tsx tests/unit/storage-config.test.ts && tsx tests/unit/checkpoint-tar.test.ts && tsx tests/unit/checkpoint-tar-listing.test.ts && tsx tests/unit/checkpoint-errors.test.ts && tsx tests/unit/checkpoint-flows.test.ts && tsx tests/unit/checkpoint-dx.test.ts && tsx tests/unit/checkpoint-edge-cases.test.ts && tsx tests/unit/codex-parser-errors.test.ts && tsx tests/unit/parser-harness-errors.test.ts && tsx tests/unit/codex-toml.test.ts && tsx tests/unit/kimi-parser.test.ts && tsx tests/unit/kimi-mcp.test.ts && tsx tests/unit/opencode-parser.test.ts && tsx tests/unit/droid-parser.test.ts && tsx tests/unit/droid-mcp.test.ts && tsx tests/unit/parser-tool-name.test.ts && tsx tests/unit/parser-atif-fields.test.ts && tsx tests/unit/provider-parity.test.ts && tsx tests/unit/storage-client.test.ts && tsx tests/unit/sessions-client.test.ts && tsx tests/unit/hosted-client.test.ts && tsx tests/unit/hosted-trial-tree.test.ts && tsx tests/unit/hosted-ergonomics.test.ts && tsx tests/unit/cli.test.ts && tsx tests/unit/cli-bin.test.ts && tsx tests/unit/cli-skills.test.ts && tsx tests/unit/cli-help.test.ts && tsx tests/unit/hosted-types.test.ts && tsx tests/unit/hosted-error-codes.test.ts && tsx tests/unit/hosted-spec-gate.test.ts && tsx tests/unit/harness-capabilities.test.ts && tsx tests/unit/sandbox-artifacts.test.ts && tsx tests/unit/upload-file-from-path.test.ts && tsx tests/unit/file-utils.test.ts && tsx tests/unit/hosted-upload-stream.test.ts && tsx tests/unit/hosted-upload-memory.test.ts && tsx tests/unit/hosted-resumable.test.ts && tsx tests/unit/cli-ndjson-pipe.test.ts && tsx tests/unit/sandbox-errors.test.ts && tsx --expose-gc tests/unit/hosted-tar.test.ts",
+    "test:unit": "node scripts/check-test-syntax.mjs && tsx tests/unit/semaphore.test.ts && tsx tests/unit/swarm-concurrency.test.ts && tsx tests/unit/swarm-verify.test.ts && tsx tests/unit/swarm-retry-verify.test.ts && tsx tests/unit/prompt-construction.test.ts && tsx tests/unit/observability-metadata.test.ts && tsx tests/unit/observability-identity.test.ts && tsx tests/unit/skills-integration.test.ts && tsx tests/unit/skills-resolver.test.ts && tsx tests/unit/skills-tag-peel.test.ts && tsx tests/unit/auth-config.test.ts && tsx tests/unit/config-validation.test.ts && tsx tests/unit/agent-presets.test.ts && tsx tests/unit/run-resume-option.test.ts && tsx tests/unit/native-config-delivery.test.ts && tsx tests/unit/browser-config.test.ts && tsx tests/unit/integrations-config.test.ts && tsx tests/unit/managed-secrets.test.ts && tsx tests/unit/managed-modal.test.ts && tsx tests/unit/plugins-config.test.ts && tsx tests/unit/session-runtime.test.ts && tsx tests/unit/cost-api.test.ts && tsx tests/unit/storage-config.test.ts && tsx tests/unit/checkpoint-tar.test.ts && tsx tests/unit/checkpoint-tar-listing.test.ts && tsx tests/unit/checkpoint-errors.test.ts && tsx tests/unit/checkpoint-flows.test.ts && tsx tests/unit/checkpoint-dx.test.ts && tsx tests/unit/checkpoint-edge-cases.test.ts && tsx tests/unit/codex-parser-errors.test.ts && tsx tests/unit/parser-harness-errors.test.ts && tsx tests/unit/codex-toml.test.ts && tsx tests/unit/kimi-parser.test.ts && tsx tests/unit/kimi-mcp.test.ts && tsx tests/unit/opencode-parser.test.ts && tsx tests/unit/droid-parser.test.ts && tsx tests/unit/droid-mcp.test.ts && tsx tests/unit/pi-parser.test.ts && tsx tests/unit/prime-agent-parser.test.ts && tsx tests/unit/pi-family-mcp.test.ts && tsx tests/unit/dsh-parser.test.ts && tsx tests/unit/dsh-mcp.test.ts && tsx tests/unit/zcode-parser.test.ts && tsx tests/unit/zcode-mcp.test.ts && tsx tests/unit/antigravity-parser.test.ts && tsx tests/unit/antigravity-mcp.test.ts && tsx tests/unit/parser-tool-name.test.ts && tsx tests/unit/parser-atif-fields.test.ts && tsx tests/unit/provider-parity.test.ts && tsx tests/unit/storage-client.test.ts && tsx tests/unit/sessions-client.test.ts && tsx tests/unit/hosted-client.test.ts && tsx tests/unit/hosted-trial-tree.test.ts && tsx tests/unit/hosted-ergonomics.test.ts && tsx tests/unit/cli.test.ts && tsx tests/unit/cli-bin.test.ts && tsx tests/unit/cli-skills.test.ts && tsx tests/unit/cli-help.test.ts && tsx tests/unit/hosted-types.test.ts && tsx tests/unit/hosted-error-codes.test.ts && tsx tests/unit/hosted-spec-gate.test.ts && tsx tests/unit/harness-capabilities.test.ts && tsx tests/unit/harness-uniformity.test.ts && tsx tests/unit/sandbox-artifacts.test.ts && tsx tests/unit/upload-file-from-path.test.ts && tsx tests/unit/file-utils.test.ts && tsx tests/unit/hosted-upload-stream.test.ts && tsx tests/unit/hosted-upload-memory.test.ts && tsx tests/unit/hosted-resumable.test.ts && tsx tests/unit/cli-ndjson-pipe.test.ts && tsx tests/unit/sandbox-errors.test.ts && tsx --expose-gc tests/unit/hosted-tar.test.ts",
     "test:unit:upload-from-path": "tsx tests/unit/upload-file-from-path.test.ts",
     "test:unit:upload-stream": "tsx tests/unit/hosted-upload-stream.test.ts",
     "test:unit:upload-memory": "tsx tests/unit/hosted-upload-memory.test.ts",
     "test:unit:parity": "tsx tests/unit/provider-parity.test.ts",
+    "test:unit:uniformity": "tsx tests/unit/harness-uniformity.test.ts",
     "test:unit:semaphore": "tsx tests/unit/semaphore.test.ts",
     "test:unit:swarm": "tsx tests/unit/swarm-concurrency.test.ts",
     "test:unit:verify": "tsx tests/unit/swarm-verify.test.ts",
@@ -89,6 +90,8 @@
     "test:unit:tool-name": "tsx tests/unit/parser-tool-name.test.ts",
     "test:unit:kimi-parser": "tsx tests/unit/kimi-parser.test.ts",
     "test:unit:kimi-mcp": "tsx tests/unit/kimi-mcp.test.ts",
+    "test:unit:dsh-parser": "tsx tests/unit/dsh-parser.test.ts",
+    "test:unit:dsh-mcp": "tsx tests/unit/dsh-mcp.test.ts",
     "test:unit:codex-toml": "tsx tests/unit/codex-toml.test.ts",
     "pretest:integration": "npm run build",
     "test:integration": "npm run test:01",
@@ -129,9 +132,9 @@
   },
   "dependencies": {
     "@agentclientprotocol/sdk": "^0.5.1",
-    "@evolvingmachines/daytona": "0.0.66",
-    "@evolvingmachines/e2b": "0.0.66",
-    "@evolvingmachines/modal": "0.0.66",
+    "@evolvingmachines/daytona": "0.0.67-train-harness-wave-2026-09-25.20260926.e00a8a6d",
+    "@evolvingmachines/e2b": "0.0.67-train-harness-wave-2026-09-25.20260926.e00a8a6d",
+    "@evolvingmachines/modal": "0.0.67-train-harness-wave-2026-09-25.20260926.e00a8a6d",
     "ajv": "^8.17.1",
     "p-map": "^7.0.2",
     "smol-toml": "^1.7.0",

package/skills/evolve-evals/references/cli-reference/analyze.md

@@ -18,7 +18,7 @@ evolve analyze --show-defaults
 evolve analyze --show-defaults --json
 ```
 
-This returns the current model, reasoning effort, sandbox provider, prompt, and rubric. Defaults mode takes no job ID or other analysis options.
+This returns the current agent, model, reasoning effort, sandbox provider, prompt, and rubric. Defaults mode takes no job ID or other analysis options.
 
 ## Select trials
 
@@ -51,6 +51,7 @@ Cancelled trials are excluded. `--passing` and `--failing` cannot be combined. T
 
 ```bash
 evolve analyze "$JOB_ID" \
+  --agent codex --model gpt-5.5 \
   --rubric rubric.toml \
   --prompt prompt.txt \
   --watch
@@ -58,7 +59,8 @@ evolve analyze "$JOB_ID" \
 
 | Option | Meaning |
 | --- | --- |
-| `-m`, `--model <name>` | Analyzer model. |
+| `-a`, `--agent <name>` | Agent the analyzer runs on. Default `claude`. |
+| `-m`, `--model <name>` | Analyzer model, from that agent's models. |
 | `--effort <value>` | Reasoning effort. |
 | `-r`, `--rubric <path>` | TOML, YAML, or JSON rubric. |
 | `-p`, `--prompt <path>` | Text file replacing the default prompt. |

package/skills/evolve-evals/references/cli-reference/check.md

@@ -44,12 +44,13 @@ evolve check --show-defaults
 | Option | Meaning |
 | --- | --- |
 | `--name <name>` | Check label. Defaults to the acceptance timestamp. |
-| `-m`, `--model <name>` | Checker model. |
+| `-a`, `--agent <name>` | Agent the checker runs on. Default `claude`. |
+| `-m`, `--model <name>` | Checker model, from that agent's models. |
 | `--effort <value>` | Reasoning effort. |
 | `-r`, `--rubric <path>` | TOML, YAML, or JSON rubric. |
 | `-p`, `--prompt <path>` | Replacement prompt text file. |
 | `-e`, `--env <provider>` | Checker sandbox provider. |
-| `--show-defaults` | Print the prompt, rubric, model, effort, and provider, then exit. |
+| `--show-defaults` | Print the agent, model, effort, provider, prompt, and rubric, then exit. |
 | `--watch` | Wait until every task check settles. |
 | `-q`, `--quiet` | With watch, suppress intermediate progress. |
 

package/skills/evolve-evals/references/cli-reference/run.md

@@ -148,7 +148,8 @@ See [Jobs](/core-concepts/jobs) for the retry and spend rules.
 | Option | Meaning |
 | --- | --- |
 | `--analyze` | Enable analysis with the platform defaults. |
-| `--analyze-model <name>` | Analyzer model. |
+| `--analyze-agent <name>` | Agent the analyzer runs on. Default `claude`. |
+| `--analyze-model <name>` | Analyzer model, from that agent's models. |
 | `--analyze-rubric <path>` | TOML, YAML, or JSON rubric. |
 | `--analyze-prompt <path>` | Replacement prompt file. |
 | `--analyze-provider <provider>` | Analyzer sandbox provider. |

package/skills/evolve-evals/references/core-concepts/agents.md

@@ -22,6 +22,11 @@ One arm
 | `kimi` | Kimi Code |
 | `opencode` | OpenCode |
 | `droid` | Droid |
+| `pi` | pi |
+| `prime-agent` | Prime Agent |
+| `dsh` | DeepSeek Harness |
+| `zcode` | Z Code |
+| `antigravity` | Antigravity CLI |
 
 ```bash
 evolve run -d harbor-examples@1.0 -i hello-world \

package/skills/evolve-evals/references/core-concepts/analyze.md

@@ -104,7 +104,7 @@ evolve analyze "$JOB_ID" -r rubric.toml -p prompt.txt --watch
 
 An analyzer prompt can use `{trial_path}`, `{task_section}`, and `{criteria_guidance}`. Evolve appends the required output format.
 
-Use `-m`, `--effort`, `-e`, and `-n` for model, effort, provider, and concurrency. On job creation, use the corresponding `--analyze-*` flags. See the [full reference](/cli-reference/analyze).
+Use `-a`, `-m`, `--effort`, `-e`, and `-n` for agent, model, effort, provider, and concurrency. The analyzer runs on `claude` unless `-a` names another agent; the model must be one of that agent's [models](/core-concepts/models#analysis-and-check-models). On job creation, use the corresponding `--analyze-*` flags. See the [full reference](/cli-reference/analyze).
 
 ## The result
 

package/skills/evolve-evals/references/core-concepts/check.md

@@ -52,13 +52,13 @@ A check is a hosted model run and incurs cost. `evolve dataset check` is a separ
 | Execution | Valid reference solution; rejection of non-solutions; environment works; stable verification; sufficient limits. |
 | Feasibility | The task is solvable. |
 
-Inspect the default rubric's criteria and their complete guidance, and the current model, effort, provider, and prompt:
+Inspect the default rubric's criteria and their complete guidance, and the current agent, model, effort, provider, and prompt:
 
 ```bash
 evolve check --show-defaults
 ```
 
-Use `-r rubric.toml` or `-p prompt.txt` to customize. Check prompt tokens are `{task_path}`, `{file_tree}`, and `{criteria_guidance}`. The required result format is appended automatically.
+Use `-r rubric.toml` or `-p prompt.txt` to customize. Check prompt tokens are `{task_path}`, `{file_tree}`, and `{criteria_guidance}`. The required result format is appended automatically. The checker runs on `claude` unless `-a` names another agent; pass a model from that agent's [models](/core-concepts/models#analysis-and-check-models) with `-m`.
 
 ## The result
 

package/skills/evolve-evals/references/core-concepts/models.md

@@ -58,15 +58,15 @@ In `/api/meta`, `agents[].effort_support` identifies the mode: `level`, `binary`
 
 ## Analysis and check models
 
-Trace analysis and task checks use the Claude Code harness. Select their model separately:
+Trace analysis and task checks run on Claude Code unless you choose another agent. Select their agent and model separately from the evaluation arms:
 
-| Operation | Model option |
-| --- | --- |
-| `evolve analyze` | `-m` |
-| `evolve check` | `-m` |
-| Analysis attached to `evolve run` | `--analyze-model` |
+| Operation | Agent option | Model option |
+| --- | --- | --- |
+| `evolve analyze` | `-a` | `-m` |
+| `evolve check` | `-a` | `-m` |
+| Analysis attached to `evolve run` | `--analyze-agent` | `--analyze-model` |
 
-Their default is `openrouter/deepseek/deepseek-v4.1-flash`. Choose from the supported Claude Code roster.
+The default model is `openrouter/deepseek/deepseek-v4.1-flash`. The model must be one of the chosen agent's models, so with an agent that does not list the default, pass `-m` too. In `/api/meta`, `analyze.agents` lists the agents these commands accept, each agent's models, and the effort each model runs at when you pass none.
 
 **[Configure a run](/cli-reference/run)**
 

package/skills/evolve-evals/references/sdk-reference/analyses.md

@@ -49,8 +49,9 @@ The field names are the same in both languages. TypeScript passes an object; Pyt
 
 | Field | Purpose |
 | --- | --- |
-| `model_name` | Analyzer model |
-| `reasoning_effort` | Effort supported by the analyzer |
+| `agent` | Agent the analyzer runs on; default `claude` |
+| `model_name` | Analyzer model, one of that agent's models |
+| `reasoning_effort` | Effort that agent accepts |
 | `sandbox_provider` | Analyzer's sandbox provider |
 | `rubric` | `{criteria: [{name, description, guidance}, ...]}` |
 | `prompt` | Replacement prompt template text |
@@ -84,7 +85,7 @@ defaults = await analyses().defaults()
 print(defaults["model_name"], defaults["rubric"], defaults["prompt"])
 ```
 
-The response includes `model_name`, `rubric`, `prompt`, `reasoning_effort`, and `sandbox_provider`. The prompt is the editable template text.
+The response includes `agent`, `model_name`, `rubric`, `prompt`, `reasoning_effort`, and `sandbox_provider`. The prompt is the editable template text.
 
 A custom analyze prompt may use `{trial_path}`, `{task_section}`, and `{criteria_guidance}`. The required result schema is appended by the platform. See [analysis concepts](/core-concepts/analyze) for rubric design.
 

package/skills/evolve-evals/references/sdk-reference/checks.md

@@ -55,8 +55,9 @@ Choose one source. A local check does not first publish a dataset.
 | Option | Purpose |
 | --- | --- |
 | `name` | Check label; defaults to an acceptance timestamp |
-| `model_name` | Checker model |
-| `reasoning_effort` | Supported reasoning effort |
+| `agent` | Agent the checker runs on; default `claude` |
+| `model_name` | Checker model, one of that agent's models |
+| `reasoning_effort` | Effort that agent accepts |
 | `sandbox_provider` | Checker sandbox provider |
 | `rubric` | `{criteria: [{name, description, guidance}, ...]}` |
 | `prompt` | Replacement prompt template |

package/skills/evolve-evals/references/sdk-reference/meta.md

@@ -34,7 +34,7 @@ Also available as `hosted().meta()` in both languages.
 | `sandbox_providers` | Which providers support the task's resources and features? |
 | `platform_constraints` | Which requirements cannot run on this deployment? |
 | `network_modes` | Which network modes can a task declare? |
-| `analyze` | Which analyzer models and reasoning efforts are available? |
+| `analyze` | Which agents can analyses and checks run on, with which models and reasoning efforts? |
 | `limits` | What are the current defaults and bounds for jobs, pages, uploads, and comparisons? |
 | `statuses` | Which states exist, and which are terminal? |
 | `agent_registration` | Which names, environment keys, and upload sizes can a custom agent use? |

package/skills/evolve-evals/references/sdk-reference/methods/checks.md

@@ -37,6 +37,7 @@ async def create(
     *,
     dataset: Optional[str] = None,
     name: Optional[str] = None,
+    agent: Optional[str] = None,
     model_name: Optional[str] = None,
     rubric: Optional[Rubric] = None,
     prompt: Optional[str] = None,
@@ -80,6 +81,7 @@ Optional `onUploadProgress(sent, total)` / `on_upload_progress(sent, total)` rep
 ```ts
 interface CheckConfigInput {
   name?: string;
+  agent?: string;
   model_name?: string;
   rubric?: Rubric;
   prompt?: string;

package/skills/evolve-evals/references/sdk-reference/methods/jobs.md

@@ -528,6 +528,7 @@ analyze(
 async def analyze(
     id: str,
     *,
+    agent: Optional[str] = None,
     model_name: Optional[str] = None,
     rubric: Optional[Rubric] = None,
     prompt: Optional[str] = None,
@@ -556,7 +557,7 @@ job = await client.analyze(
 )
 ```
 
-All options are optional. `analyses().defaults()` reads the current model, effort, provider, rubric, and prompt defaults.
+All options are optional. `analyses().defaults()` reads the current agent, model, effort, provider, rubric, and prompt defaults.
 
 | Selection | Meaning |
 | --- | --- |
@@ -572,6 +573,7 @@ Cancelled trials are excluded. `trial_ids` is accepted here, but refused inside
 
 ```ts
 interface AnalyzeConfigInput {
+  agent?: string;
   model_name?: string;
   rubric?: Rubric;
   prompt?: string;

package/skills/evolve-evals/references/sdk-reference/methods/meta.md

@@ -77,9 +77,13 @@ interface CapabilityDocument {
     dataset_version: StatusVocabulary;
   };
   analyze?: {
+    default_agent: string;
     default_model: string;
-    reasoning_efforts: string[];
-    models: { alias: string; model_id: string; default_reasoning_effort: string }[];
+    agents: {
+      name: string;
+      reasoning_efforts: string[];
+      models: { alias: string; model_id: string; default_reasoning_effort: string | null }[];
+    }[];
   };
   limits: {
     job: {