opencode-agile-agent 1.0.4 → 1.2.1

package/templates/.opencode/commands/assign-models.md

@@ -0,0 +1,39 @@
+---
+description: Recommend and assign the best available models to each agent.
+agent: feature-lead
+---
+
+Ask the user to paste the output of `opencode models` before doing any mapping work. If they have not provided it yet, stop and ask for it.
+
+After the user provides the model list:
+
+1. Inspect `.opencode/agents/*.md` plus the active OpenCode config file.
+2. Prefer updating `opencode.json` when it exists. If missing, create `opencode.json` at the project root.
+3. Parse the available models into practical tiers based on reasoning strength, context/token capacity, latency/cost, and likely tool reliability.
+4. Score each agent by:
+   - reasoning depth
+   - need for long context/token budget
+   - risk if the agent makes a bad call
+   - creativity/exploration needs
+   - strictness implied by the current `temperature` and `top_p`
+5. Produce a proposed mapping for every agent, sorted from highest-priority model assignments to lowest-priority assignments.
+6. For each assignment, explain why that model fits the agent better than lighter or heavier options.
+
+Guidance for matching:
+
+- Use the strongest reasoning models for orchestration, architecture, debugging, security, and high-risk review agents when those models are available.
+- Use balanced mid-tier models for implementation specialists and domain experts that need solid reasoning but do not justify the most expensive model every time.
+- Use lightweight/fast models for low-risk scouting, inventory, or support work when quality tradeoffs are acceptable.
+- Treat lower `temperature` and lower `top_p` as a signal that the agent benefits from stable, deterministic behavior. Do not change those values unless the user asks.
+- Treat higher-context models as more valuable for agents that synthesize many files, plans, or reviews.
+- If two models are close, prefer the cheaper/faster one for lower-impact agents.
+- If the user only provided a small model list, reuse models where needed and call out the tradeoff explicitly.
+
+Before editing anything, ask for confirmation on the proposed mapping.
+
+After approval:
+
+1. Update `opencode.json`.
+2. Preserve existing agent settings unless they conflict with the approved model mapping.
+3. Add missing agent entries when needed, using the agent frontmatter defaults for `mode`, `temperature`, `top_p`, and `steps`, then set the approved `model`.
+4. Return a concise summary of which file was updated, which agents changed, and any agents that still share a fallback model because the provided model catalog was too small.

package/templates/.opencode/commands/brainstorm.md

@@ -3,8 +3,11 @@ description: Explore options in a back-and-forth discussion before planning.
 agent: feature-lead
 ---
 
+Call `session_artifact_current` first.
+If there is no active feature yet, choose a short feature slug and treat this stage as candidate state only.
 Start with @context-gatherer.
-Use @product-manager to frame the problem.
 Use @project-planner and @explorer-agent to gather constraints and options.
 Ask the involved agents to present the best default, the tradeoffs, and the open questions.
-Return one recommended path and the compact context bundle: proposal.md, goal.md, spec.md, task.md, and important.md.
+Do not create the full planning bundle here unless the user explicitly asks for it.
+Before finishing, call `session_artifact_update` to record the recommended path, open questions, next step, and stage `brainstorm`.
+Return one recommended path, the key tradeoffs, the open questions, and the clearest next step into `plan`.

package/templates/.opencode/commands/check-progress.md

@@ -0,0 +1,21 @@
+---
+description: Check current progress, changed files, remaining work, and the best next step from status.yaml plus git-aware evidence.
+agent: feature-lead
+---
+
+Start with @context-gatherer.
+Call `session_artifact_current` first.
+Read `status.yaml` first as the live execution source of truth.
+Treat `status.yaml.status` as one of: `brainstorm`, `planning`, `implementation`, `verification`, `review`, `done`, `blocked`.
+Then inspect the current working state and reconcile `status.yaml` with git-aware evidence: status, diff summary, staged vs unstaged changes, and recent commits if useful.
+If `status.yaml` and repo evidence disagree, call out the mismatch explicitly.
+Summarize the work in this format:
+- Current status
+- Done
+- Changed
+- In progress
+- Remaining
+- Risks or blockers
+- Next best step
+Keep it concrete and decision-ready.
+Do not give a generic status report; tie each point to visible evidence in the repo.

package/templates/.opencode/commands/create.md

@@ -3,9 +3,14 @@ description: Build new features, components, or project slices with a spec-drive
 agent: feature-lead
 ---
 
+Call `session_artifact_current` first.
+Call `session_artifact_handoff` for `developer` before delegating.
+Call `session_artifact_acceptance_criteria` when the implementation target needs a tight contract.
 Start with @context-gatherer.
-Use @system-analyst to refresh the compact context bundle: proposal.md, goal.md, spec.md, task.md, and important.md.
+Read the approved compact context bundle first: brief.md, spec.md, task.md, notes.md, and status.yaml.
+If the bundle is missing or materially unclear, stop and use `plan` before implementation.
 Hand implementation to @developer.
-If the work touches sensitive paths, add @security-auditor and @penetration-tester before approval.
+Call `session_artifact_update` to set stage `implementation`, the current owner, files of interest, and next step.
+If the work touches sensitive paths, add @security-auditor before approval.
 Close the loop with @test-engineer and @pr-reviewer.
-Archive the approved bundle in `.opencode/archive/<feature-slug>/`.
+If the work is approved, use the `archive` command to preserve the finished bundle.

package/templates/.opencode/commands/plan.md

@@ -3,7 +3,12 @@ description: Create structured task breakdowns and spec artifacts.
 agent: feature-lead
 ---
 
+Call `session_artifact_current` first.
+Promote the active feature into planning state if it is still in brainstorm state.
 Start with @context-gatherer.
+Call `session_artifact_handoff` for `system-analyst` before delegating.
 Use @project-planner to break the work into atomic tasks.
-Use @system-analyst to write proposal.md, goal.md, spec.md, task.md, and important.md.
-Return the ordered plan, dependencies, and exit criteria.
+Use @system-analyst to write brief.md, spec.md, task.md, notes.md, and status.yaml.
+Call `session_artifact_update` to set stage `planning`, owner, handoff target, approved scope, risks, and next step.
+If the `submit_plan` tool is available, call it with the final plan markdown before returning. If the plan is denied, incorporate the feedback and resubmit.
+Return the ordered plan, dependencies, exit criteria, and the approval result (approved/changes requested).

package/templates/.opencode/commands/reframe.md

@@ -0,0 +1,17 @@
+---
+description: Reset the framing when the output is off-target, unclear, or stuck repeating the same mistake.
+agent: feature-lead
+---
+
+Do not jump straight into another fix.
+First restate what you think the user actually wants in plain language.
+Then explain what seems mismatched between expected and actual output.
+List:
+- current understanding
+- likely misunderstanding
+- what should be kept
+- what should be discarded
+- 2-3 better interpretations or directions
+- the clearest next step
+If the request is still ambiguous, ask short targeted clarifying questions.
+Prefer correcting the frame over repeating the same implementation pattern.

package/templates/.opencode/commands/review.md

@@ -3,9 +3,15 @@ description: Review code against the spec, standards, and quality gates.
 agent: feature-lead
 ---
 
+Call `session_artifact_current` first.
+Call `session_artifact_handoff` for `pr-reviewer` before delegating.
+Call `session_artifact_review_packet` before review so the reviewer can compare code against the approved scope without reloading the full bundle.
+Call `session_artifact_repo_delta` before review and flag any mismatch between artifact state and git state.
 Start with @context-gatherer.
-If the change touches sensitive paths, insert @security-auditor first and use @penetration-tester for the redteam phase when needed.
-Hand the change to @pr-reviewer.
+Read the approved bundle and current `status.yaml` first.
+If the change touches sensitive paths, insert @security-auditor first.
+Hand the change to @pr-reviewer to compare implementation against the approved spec, task scope, and quality gates.
 If changes are requested, loop back to @developer.
-Archive the approved bundle in `.opencode/archive/<feature-slug>/` after approval.
+Call `session_artifact_update` to set stage `review` and record the review outcome, next owner, and next step.
+If the change is approved, recommend the `archive` command as the next step.
 Return APPROVED or CHANGES REQUESTED with actionable feedback.

package/templates/.opencode/commands/rubber-duck.md

@@ -0,0 +1,14 @@
+---
+description: Think out loud, challenge assumptions, and isolate the real problem before planning or editing.
+agent: feature-lead
+---
+
+Start with @context-gatherer only if current project state matters.
+Use a rubber-duck style conversation before proposing changes.
+First restate the problem in plain language.
+Then list the current facts, assumptions, unknowns, and contradictions.
+Ask short clarifying questions if needed.
+Generate 2-4 plausible explanations for why the issue or request is still unclear or unresolved.
+For each explanation, give the best next check.
+Do not rush into implementation.
+Return the most likely current understanding, the key uncertainty, and the next best step.

package/templates/.opencode/commands/status.md

@@ -3,7 +3,10 @@ description: Check project health, docs, and operating readiness.
 agent: feature-lead
 ---
 
+Call `session_artifact_current` first.
+Call `session_artifact_repo_delta` before summarizing status.
 Start with @context-gatherer.
 Use @explorer-agent to map the repo and conventions.
 Use @test-engineer and @devops-engineer when quality signals are needed.
+Reconcile the artifact state with repo evidence. If they disagree, call out the mismatch explicitly.
 Summarize the health, risks, missing docs, and next step.

package/templates/.opencode/commands/test.md

@@ -3,8 +3,13 @@ description: Run and generate tests for the codebase.
 agent: feature-lead
 ---
 
+Call `session_artifact_current` first.
+Call `session_artifact_handoff` for `test-engineer` before delegating.
+Call `session_artifact_changed_files` and `session_artifact_acceptance_criteria` before choosing test scope.
 Start with @context-gatherer.
-Use @test-engineer to run or write the right test layer.
-Use @qa-automation-engineer when the harness needs work.
+Read the approved bundle and current `status.yaml` first.
+Use @test-engineer to verify the acceptance criteria, changed behavior, and highest-risk nearby paths.
+Load the `cross-model-regression` skill when the user wants a multi-agent regression sweep or cross-model verification.
 If a bug surfaces, hand the failing case to @developer.
-Return the important paths, coverage gaps, and the next verification step.
+Call `session_artifact_update` to set stage `verification`, capture verified scope, failing paths, and the next verification step.
+Return what was verified, any coverage gaps, failing paths, and the next verification step.

package/templates/.opencode/config.template.json

@@ -1,5 +1,5 @@
 {
-  "$schema": "https://json.schemastore.org/opencode-config.json",
+  "$schema": "https://opencode.ai/config.json",
   "version": "1.0.0",
   "project": {
     "name": "Your Project Name",
@@ -8,54 +8,192 @@
     "language": "typescript|javascript",
     "styling": "tailwind|css-modules|styled-components|scss|other"
   },
-  "agents": {
-    "featureLead": {
+  "agent": {
+    "feature-lead": {
       "enabled": true,
       "model": "primary",
-      "mode": "primary"
+      "mode": "primary",
+      "temperature": 0.3,
+      "top_p": 0.9,
+      "steps": 50,
+      "permission": {
+        "task": {
+          "*": "allow",
+          "security-auditor": "ask"
+        }
+      }
     },
-    "systemAnalyst": {
+    "system-analyst": {
       "enabled": true,
-      "model": "opencode-go/kimi-k2.5",
       "mode": "subagent",
+      "temperature": 0.1,
+      "top_p": 0.85,
+      "steps": 30,
+      "permission": {
+        "task": {
+          "*": "deny",
+          "explorer-agent": "allow",
+          "context-gatherer": "allow"
+        }
+      },
       "tools": {
         "write": true,
         "edit": true,
         "bash": false
       }
     },
+    "context-gatherer": {
+      "enabled": true,
+      "mode": "subagent",
+      "temperature": 0.1,
+      "top_p": 0.8,
+      "steps": 20
+    },
+    "project-planner": {
+      "enabled": true,
+      "mode": "subagent",
+      "temperature": 0.2,
+      "top_p": 0.85,
+      "steps": 35
+    },
+    "explorer-agent": {
+      "enabled": true,
+      "mode": "subagent",
+      "temperature": 0.1,
+      "top_p": 0.75,
+      "steps": 25
+    },
+    "archiver": {
+      "enabled": true,
+      "mode": "subagent",
+      "temperature": 0.1,
+      "top_p": 0.82,
+      "steps": 30
+    },
     "developer": {
       "enabled": true,
-      "model": "github-copilot/gpt-5.1-codex-mini",
       "mode": "subagent",
+      "temperature": 0.2,
+      "top_p": 0.9,
+      "steps": 100,
+      "permission": {
+        "task": {
+          "*": "ask",
+          "pr-reviewer": "allow",
+          "test-engineer": "allow"
+        }
+      },
       "tools": {
         "write": true,
         "edit": true,
         "bash": true
       }
     },
-    "prReviewer": {
+    "backend-specialist": {
       "enabled": true,
-      "model": "github-copilot/claude-haiku-4.5",
       "mode": "subagent",
+      "temperature": 0.25,
+      "top_p": 0.9,
+      "steps": 80
+    },
+    "frontend-specialist": {
+      "enabled": true,
+      "mode": "subagent",
+      "temperature": 0.25,
+      "top_p": 0.9,
+      "steps": 70
+    },
+    "devops-engineer": {
+      "enabled": true,
+      "mode": "subagent",
+      "temperature": 0.2,
+      "top_p": 0.85,
+      "steps": 50
+    },
+    "test-engineer": {
+      "enabled": true,
+      "mode": "subagent",
+      "temperature": 0.15,
+      "top_p": 0.82,
+      "steps": 60,
+      "permission": {
+        "task": {
+          "*": "deny"
+        }
+      },
+      "tools": {
+        "write": true,
+        "edit": true,
+        "bash": true
+      }
+    },
+    "pr-reviewer": {
+      "enabled": true,
+      "mode": "subagent",
+      "temperature": 0.1,
+      "top_p": 0.8,
+      "steps": 40,
+      "permission": {
+        "task": {
+          "*": "deny"
+        }
+      },
       "tools": {
         "write": false,
         "edit": false,
         "bash": false
       }
+    },
+    "security-auditor": {
+      "enabled": true,
+      "mode": "subagent",
+      "temperature": 0.1,
+      "top_p": 0.85,
+      "steps": 45
+    },
+    "debugger": {
+      "enabled": true,
+      "model": "primary",
+      "mode": "subagent",
+      "temperature": 0.15,
+      "top_p": 0.85,
+      "steps": 60,
+      "permission": {
+        "task": {
+          "*": "deny"
+        }
+      },
+      "tools": {
+        "write": true,
+        "edit": true,
+        "bash": true
+      }
+    },
+    "performance-optimizer": {
+      "enabled": true,
+      "mode": "subagent",
+      "temperature": 0.2,
+      "top_p": 0.88,
+      "steps": 60,
+      "permission": {
+        "task": {
+          "*": "deny"
+        }
+      },
+      "tools": {
+        "write": true,
+        "edit": true,
+        "bash": true
+      }
     }
   },
-  "ralph": {
-    "enabled": true,
-    "defaultMaxIterations": 10,
-    "defaultTimeoutMs": 300000,
-    "parallelDefault": false,
-    "autoVerify": true,
-    "exportCheckpoints": true,
-    "validatePrd": true,
-    "detectEdgeCases": true
-  },
   "plugins": {
+    "sessionArtifacts": {
+      "enabled": true,
+      "candidateStage": "brainstorm",
+      "archiveCommand": "archive",
+      "handoffTool": "session_artifact_handoff"
+    },
     "lintFormat": {
       "enabled": true,
       "runOnEdit": true,
@@ -117,9 +255,11 @@
     "agents": ".opencode/agents",
     "skills": ".opencode/skills",
     "plugins": ".opencode/plugins",
+    "artifacts": ".opencode/artifacts",
     "templates": ".opencode/templates",
     "context": "context",
-    "output": "output"
+    "output": "output",
+    "checkpoints": ".opencode/checkpoints"
   },
   "logging": {
     "level": "info",

package/templates/.opencode/package-lock.json

@@ -0,0 +1,115 @@
+{
+  "name": ".opencode",
+  "lockfileVersion": 3,
+  "requires": true,
+  "packages": {
+    "": {
+      "dependencies": {
+        "@opencode-ai/plugin": "1.4.3"
+      }
+    },
+    "node_modules/@opencode-ai/plugin": {
+      "version": "1.4.3",
+      "resolved": "https://registry.npmjs.org/@opencode-ai/plugin/-/plugin-1.4.3.tgz",
+      "integrity": "sha512-Ob/3tVSIeuMRJBr2O23RtrnC5djRe01Lglx+TwGEmjrH9yDBJ2tftegYLnNEjRoMuzITgq9LD8168p4pzv+U/A==",
+      "license": "MIT",
+      "dependencies": {
+        "@opencode-ai/sdk": "1.4.3",
+        "zod": "4.1.8"
+      },
+      "peerDependencies": {
+        "@opentui/core": ">=0.1.97",
+        "@opentui/solid": ">=0.1.97"
+      },
+      "peerDependenciesMeta": {
+        "@opentui/core": {
+          "optional": true
+        },
+        "@opentui/solid": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@opencode-ai/sdk": {
+      "version": "1.4.3",
+      "resolved": "https://registry.npmjs.org/@opencode-ai/sdk/-/sdk-1.4.3.tgz",
+      "integrity": "sha512-X0CAVbwoGAjTY2iecpWkx2B+GAa2jSaQKYpJ+xILopeF/OGKZUN15mjqci+L7cEuwLHV5wk3x2TStUOVCa5p0A==",
+      "license": "MIT",
+      "dependencies": {
+        "cross-spawn": "7.0.6"
+      }
+    },
+    "node_modules/cross-spawn": {
+      "version": "7.0.6",
+      "resolved": "https://registry.npmjs.org/cross-spawn/-/cross-spawn-7.0.6.tgz",
+      "integrity": "sha512-uV2QOWP2nWzsy2aMp8aRibhi9dlzF5Hgh5SHaB9OiTGEyDTiJJyx0uy51QXdyWbtAHNua4XJzUKca3OzKUd3vA==",
+      "license": "MIT",
+      "dependencies": {
+        "path-key": "^3.1.0",
+        "shebang-command": "^2.0.0",
+        "which": "^2.0.1"
+      },
+      "engines": {
+        "node": ">= 8"
+      }
+    },
+    "node_modules/isexe": {
+      "version": "2.0.0",
+      "resolved": "https://registry.npmjs.org/isexe/-/isexe-2.0.0.tgz",
+      "integrity": "sha512-RHxMLp9lnKHGHRng9QFhRCMbYAcVpn69smSGcq3f36xjgVVWThj4qqLbTLlq7Ssj8B+fIQ1EuCEGI2lKsyQeIw==",
+      "license": "ISC"
+    },
+    "node_modules/path-key": {
+      "version": "3.1.1",
+      "resolved": "https://registry.npmjs.org/path-key/-/path-key-3.1.1.tgz",
+      "integrity": "sha512-ojmeN0qd+y0jszEtoY48r0Peq5dwMEkIlCOu6Q5f41lfkswXuKtYrhgoTpLnyIcHm24Uhqx+5Tqm2InSwLhE6Q==",
+      "license": "MIT",
+      "engines": {
+        "node": ">=8"
+      }
+    },
+    "node_modules/shebang-command": {
+      "version": "2.0.0",
+      "resolved": "https://registry.npmjs.org/shebang-command/-/shebang-command-2.0.0.tgz",
+      "integrity": "sha512-kHxr2zZpYtdmrN1qDjrrX/Z1rR1kG8Dx+gkpK1G4eXmvXswmcE1hTWBWYUzlraYw1/yZp6YuDY77YtvbN0dmDA==",
+      "license": "MIT",
+      "dependencies": {
+        "shebang-regex": "^3.0.0"
+      },
+      "engines": {
+        "node": ">=8"
+      }
+    },
+    "node_modules/shebang-regex": {
+      "version": "3.0.0",
+      "resolved": "https://registry.npmjs.org/shebang-regex/-/shebang-regex-3.0.0.tgz",
+      "integrity": "sha512-7++dFhtcx3353uBaq8DDR4NuxBetBzC7ZQOhmTQInHEd6bSrXdiEyzCvG07Z44UYdLShWUyXt5M/yhz8ekcb1A==",
+      "license": "MIT",
+      "engines": {
+        "node": ">=8"
+      }
+    },
+    "node_modules/which": {
+      "version": "2.0.2",
+      "resolved": "https://registry.npmjs.org/which/-/which-2.0.2.tgz",
+      "integrity": "sha512-BLI3Tl1TW3Pvl70l3yq3Y64i+awpwXqsGBYWkkqMtnbXgrMD+yj7rhW0kuEDxzJaYXGjEW5ogapKNMEKNMjibA==",
+      "license": "ISC",
+      "dependencies": {
+        "isexe": "^2.0.0"
+      },
+      "bin": {
+        "node-which": "bin/node-which"
+      },
+      "engines": {
+        "node": ">= 8"
+      }
+    },
+    "node_modules/zod": {
+      "version": "4.1.8",
+      "license": "MIT",
+      "funding": {
+        "url": "https://github.com/sponsors/colinhacks"
+      }
+    }
+  }
+}

package/templates/.opencode/package.json

@@ -0,0 +1,6 @@
+{
+  "dependencies": {
+    "@opencode-ai/plugin": "1.4.3",
+    "yaml": "^2.8.0"
+  }
+}