@harness-engineering/cli 1.1.0 → 1.2.0

package/dist/agents/{skills/gemini-cli/harness-code-review/SKILL.md → commands/gemini-cli/harness/code-review.toml}

@@ -1,3 +1,17 @@
+# Generated by harness generate-slash-commands. Do not edit.
+description = "Structured code review with automated harness checks"
+prompt = """
+<context>
+Cognitive mode: adversarial-reviewer
+Type: rigid
+</context>
+
+<objective>
+Structured code review with automated harness checks
+</objective>
+
+<execution_context>
+--- SKILL.md (agents/skills/claude-code/harness-code-review/SKILL.md) ---
 # Harness Code Review
 
 > Full code review lifecycle — request, perform, respond — with automated harness checks and technical rigor over social performance.
@@ -401,3 +415,47 @@ After addressing all feedback, re-request review with:
 - **When you cannot reproduce a reported issue:** Ask the reviewer for exact reproduction steps. If they cannot provide them, the issue may not be real.
 - **When review is taking more than 2 rounds:** If the same code is going through a third round of review, something is fundamentally misaligned. Stop and discuss the approach in a meeting or synchronous conversation.
 - **When harness checks fail and you believe the check is wrong:** Do not override or skip the check. File an issue against the harness configuration and work around the limitation until it is resolved.
+
+
+--- skill.yaml (agents/skills/claude-code/harness-code-review/skill.yaml) ---
+name: harness-code-review
+version: "1.0.0"
+description: Structured code review with automated harness checks
+cognitive_mode: adversarial-reviewer
+triggers:
+  - manual
+  - on_pr
+  - on_review
+platforms:
+  - claude-code
+  - gemini-cli
+tools:
+  - Bash
+  - Read
+  - Glob
+  - Grep
+cli:
+  command: harness skill run harness-code-review
+  args:
+    - name: path
+      description: Project root path
+      required: false
+mcp:
+  tool: run_skill
+  input:
+    skill: harness-code-review
+    path: string
+type: rigid
+state:
+  persistent: false
+  files: []
+depends_on: []
+
+</execution_context>
+
+<process>
+1. Try: invoke mcp__harness__run_skill with skill: "harness-code-review"
+2. If MCP unavailable: follow the SKILL.md workflow provided above directly
+3. Pass through any arguments provided by the user
+</process>
+"""

package/dist/agents/{agents/skills/claude-code/harness-debugging/SKILL.md → commands/gemini-cli/harness/debugging.toml}

@@ -1,3 +1,24 @@
+# Generated by harness generate-slash-commands. Do not edit.
+description = "Systematic debugging with harness validation and state tracking"
+prompt = """
+<context>
+Cognitive mode: diagnostic-investigator
+Type: rigid
+State: persistent (files: .harness/debug/)
+</context>
+
+<objective>
+Systematic debugging with harness validation and state tracking
+
+Phases:
+- investigate: Entropy analysis and root cause search
+- analyze: Pattern matching against codebase
+- hypothesize: Form and test single hypothesis (optional)
+- fix: TDD-style regression test and fix
+</objective>
+
+<execution_context>
+--- SKILL.md (agents/skills/claude-code/harness-debugging/SKILL.md) ---
 # Harness Debugging
 
 > 4-phase systematic debugging with entropy analysis and persistent sessions. Phase 1 before ANY fix. "It's probably X" is not a diagnosis.
@@ -354,3 +375,62 @@ Revert test: Commenting out the validation check causes the test to fail with 50
 - **Bug is in a dependency you do not control:** Document the bug, write a test that demonstrates it, and escalate. If a workaround is needed, clearly mark it as a workaround with a reference to the upstream issue.
 - **Investigation reveals a systemic issue:** If the bug is a symptom of a larger architectural problem (e.g., widespread race conditions, fundamental type unsafety), escalate to the human. A local fix will not solve a systemic problem.
 - **Debug session exceeds 60 minutes without progress:** Something is wrong with the approach. Stop. Summarize what you know in the session file. Take a break (context reset). Return with fresh eyes and re-read the session file from the beginning.
+
+
+--- skill.yaml (agents/skills/claude-code/harness-debugging/skill.yaml) ---
+name: harness-debugging
+version: "1.0.0"
+description: Systematic debugging with harness validation and state tracking
+cognitive_mode: diagnostic-investigator
+triggers:
+  - manual
+  - on_bug_fix
+platforms:
+  - claude-code
+  - gemini-cli
+tools:
+  - Bash
+  - Read
+  - Write
+  - Edit
+  - Glob
+  - Grep
+cli:
+  command: harness skill run harness-debugging
+  args:
+    - name: path
+      description: Project root path
+      required: false
+mcp:
+  tool: run_skill
+  input:
+    skill: harness-debugging
+    path: string
+type: rigid
+phases:
+  - name: investigate
+    description: Entropy analysis and root cause search
+    required: true
+  - name: analyze
+    description: Pattern matching against codebase
+    required: true
+  - name: hypothesize
+    description: Form and test single hypothesis
+    required: false
+  - name: fix
+    description: TDD-style regression test and fix
+    required: true
+state:
+  persistent: true
+  files:
+    - .harness/debug/
+depends_on: []
+
+</execution_context>
+
+<process>
+1. Try: invoke mcp__harness__run_skill with skill: "harness-debugging"
+2. If MCP unavailable: follow the SKILL.md workflow provided above directly
+3. Pass through any arguments provided by the user
+</process>
+"""

package/dist/agents/{agents/skills/gemini-cli/detect-doc-drift/SKILL.md → commands/gemini-cli/harness/detect-doc-drift.toml}

@@ -1,3 +1,17 @@
+# Generated by harness generate-slash-commands. Do not edit.
+description = "Detect documentation that has drifted from code"
+prompt = """
+<context>
+Cognitive mode: diagnostic-investigator
+Type: flexible
+</context>
+
+<objective>
+Detect documentation that has drifted from code
+</objective>
+
+<execution_context>
+--- SKILL.md (agents/skills/claude-code/detect-doc-drift/SKILL.md) ---
 # Detect Doc Drift
 
 > Detect documentation that has drifted from code. Find stale docs before they mislead developers and AI agents.
@@ -157,3 +171,45 @@ GAP: Undocumented module
 - **When you cannot determine the correct replacement text:** The code change may have been complex. Check the commit message and PR description for context. If still unclear, flag the finding for the original author to resolve.
 - **When documentation is in a format you cannot parse:** Some docs may be in wiki pages, Confluence, or other external systems. Report the finding with a link and flag it for manual review.
 - **When drift reveals a deeper problem (code changed but nobody knew):** This suggests a process gap. Recommend adding `harness check-docs` to the CI pipeline or pre-merge hooks to catch drift at the source.
+
+
+--- skill.yaml (agents/skills/claude-code/detect-doc-drift/skill.yaml) ---
+name: detect-doc-drift
+version: "1.0.0"
+description: Detect documentation that has drifted from code
+cognitive_mode: diagnostic-investigator
+triggers:
+  - manual
+  - on_pr
+platforms:
+  - claude-code
+  - gemini-cli
+tools:
+  - Bash
+  - Read
+  - Glob
+cli:
+  command: harness skill run detect-doc-drift
+  args:
+    - name: path
+      description: Project root path
+      required: false
+mcp:
+  tool: run_skill
+  input:
+    skill: detect-doc-drift
+    path: string
+type: flexible
+state:
+  persistent: false
+  files: []
+depends_on: []
+
+</execution_context>
+
+<process>
+1. Try: invoke mcp__harness__run_skill with skill: "detect-doc-drift"
+2. If MCP unavailable: follow the SKILL.md workflow provided above directly
+3. Pass through any arguments provided by the user
+</process>
+"""

package/dist/agents/{agents/skills/gemini-cli/harness-diagnostics/SKILL.md → commands/gemini-cli/harness/diagnostics.toml}

@@ -1,3 +1,24 @@
+# Generated by harness generate-slash-commands. Do not edit.
+description = "Classify errors into taxonomy categories and route to resolution strategies"
+prompt = """
+<context>
+Cognitive mode: diagnostic-investigator
+Type: rigid
+State: persistent (files: .harness/diagnostics/)
+</context>
+
+<objective>
+Classify errors into taxonomy categories and route to resolution strategies
+
+Phases:
+- classify: Categorize the error into one of 7 taxonomy categories
+- route: Apply the resolution strategy for the error category
+- resolve: Execute the resolution and verify the fix
+- record: Record findings in anti-pattern log if initial approach failed (optional)
+</objective>
+
+<execution_context>
+--- SKILL.md (agents/skills/claude-code/harness-diagnostics/SKILL.md) ---
 # Harness Diagnostics
 
 > Cognitive mode: **diagnostic-investigator**. Classify errors into taxonomy categories and route to deterministic resolution strategies. Evidence first, classification second, action third.
@@ -316,3 +337,65 @@ Verification: 0 type errors (unchanged), 20/20 test runs pass. Fix confirmed.
 
 **Phase 4 — RECORD:**
 Not needed — initial classification was correct and first fix attempt succeeded.
+
+
+--- skill.yaml (agents/skills/claude-code/harness-diagnostics/skill.yaml) ---
+name: harness-diagnostics
+version: "1.0.0"
+description: Classify errors into taxonomy categories and route to resolution strategies
+cognitive_mode: diagnostic-investigator
+triggers:
+  - manual
+  - on_bug_fix
+platforms:
+  - claude-code
+  - gemini-cli
+tools:
+  - Bash
+  - Read
+  - Glob
+  - Grep
+  - Edit
+  - Write
+cli:
+  command: harness skill run harness-diagnostics
+  args:
+    - name: path
+      description: Project root path
+      required: false
+    - name: error
+      description: Error message or description to diagnose
+      required: false
+mcp:
+  tool: run_skill
+  input:
+    skill: harness-diagnostics
+    path: string
+type: rigid
+phases:
+  - name: classify
+    description: Categorize the error into one of 7 taxonomy categories
+    required: true
+  - name: route
+    description: Apply the resolution strategy for the error category
+    required: true
+  - name: resolve
+    description: Execute the resolution and verify the fix
+    required: true
+  - name: record
+    description: Record findings in anti-pattern log if initial approach failed
+    required: false
+state:
+  persistent: true
+  files:
+    - .harness/diagnostics/
+depends_on: []
+
+</execution_context>
+
+<process>
+1. Try: invoke mcp__harness__run_skill with skill: "harness-diagnostics"
+2. If MCP unavailable: follow the SKILL.md workflow provided above directly
+3. Pass through any arguments provided by the user
+</process>
+"""

package/dist/agents/{agents/skills/gemini-cli/enforce-architecture/SKILL.md → commands/gemini-cli/harness/enforce-architecture.toml}

@@ -1,3 +1,17 @@
+# Generated by harness generate-slash-commands. Do not edit.
+description = "Validate architectural layer boundaries and detect circular dependencies"
+prompt = """
+<context>
+Cognitive mode: meticulous-verifier
+Type: rigid
+</context>
+
+<objective>
+Validate architectural layer boundaries and detect circular dependencies
+</objective>
+
+<execution_context>
+--- SKILL.md (agents/skills/claude-code/enforce-architecture/SKILL.md) ---
 # Enforce Architecture
 
 > Validate architectural layer boundaries and detect dependency violations. No code may violate layer constraints — this is a hard gate, not a suggestion.
@@ -163,3 +177,46 @@ These are hard stops. Architecture violations are not warnings — they are erro
 - **When `harness check-deps` reports false positives:** Verify the layer assignments in `harness.config.json` are correct. If a file is assigned to the wrong layer, fix the config. If the tool is genuinely wrong, report the issue.
 - **When fixing one violation creates another:** This usually indicates a deeper structural issue. Step back and look at the dependency graph as a whole rather than fixing violations one at a time.
 - **When the team wants to change the layer model:** This is a significant architectural decision. All existing code must be migrated to the new model. Plan this as a dedicated refactoring effort, not a side task.
+
+
+--- skill.yaml (agents/skills/claude-code/enforce-architecture/skill.yaml) ---
+name: enforce-architecture
+version: "1.0.0"
+description: Validate architectural layer boundaries and detect circular dependencies
+cognitive_mode: meticulous-verifier
+triggers:
+  - manual
+  - on_pr
+  - on_commit
+platforms:
+  - claude-code
+  - gemini-cli
+tools:
+  - Bash
+  - Read
+  - Glob
+cli:
+  command: harness skill run enforce-architecture
+  args:
+    - name: path
+      description: Project root path
+      required: false
+mcp:
+  tool: run_skill
+  input:
+    skill: enforce-architecture
+    path: string
+type: rigid
+state:
+  persistent: false
+  files: []
+depends_on: []
+
+</execution_context>
+
+<process>
+1. Try: invoke mcp__harness__run_skill with skill: "enforce-architecture"
+2. If MCP unavailable: follow the SKILL.md workflow provided above directly
+3. Pass through any arguments provided by the user
+</process>
+"""

package/dist/agents/{agents/skills/gemini-cli/harness-execution/SKILL.md → commands/gemini-cli/harness/execution.toml}

@@ -1,3 +1,24 @@
+# Generated by harness generate-slash-commands. Do not edit.
+description = "Execute a planned set of tasks with harness validation and state tracking"
+prompt = """
+<context>
+Cognitive mode: meticulous-implementer
+Type: rigid
+State: persistent (files: .harness/state.json, .harness/learnings.md)
+</context>
+
+<objective>
+Execute a planned set of tasks with harness validation and state tracking
+
+Phases:
+- prepare: Load state and verify prerequisites
+- execute: Implement planned tasks
+- verify: Validate each completed task
+- persist: Save progress to state files
+</objective>
+
+<execution_context>
+--- SKILL.md (agents/skills/claude-code/harness-execution/SKILL.md) ---
 # Harness Execution
 
 > Execute a plan task by task with atomic commits, checkpoint protocol, and persistent knowledge capture. Stop on blockers. Do not guess.
@@ -296,3 +317,65 @@ Example:
 ```
 
 This is for human debugging only. Not required for normal execution.
+
+
+--- skill.yaml (agents/skills/claude-code/harness-execution/skill.yaml) ---
+name: harness-execution
+version: "1.0.0"
+description: Execute a planned set of tasks with harness validation and state tracking
+cognitive_mode: meticulous-implementer
+triggers:
+  - manual
+  - on_new_feature
+  - on_bug_fix
+platforms:
+  - claude-code
+  - gemini-cli
+tools:
+  - Bash
+  - Read
+  - Write
+  - Edit
+  - Glob
+  - Grep
+cli:
+  command: harness skill run harness-execution
+  args:
+    - name: path
+      description: Project root path
+      required: false
+mcp:
+  tool: run_skill
+  input:
+    skill: harness-execution
+    path: string
+type: rigid
+phases:
+  - name: prepare
+    description: Load state and verify prerequisites
+    required: true
+  - name: execute
+    description: Implement planned tasks
+    required: true
+  - name: verify
+    description: Validate each completed task
+    required: true
+  - name: persist
+    description: Save progress to state files
+    required: true
+state:
+  persistent: true
+  files:
+    - .harness/state.json
+    - .harness/learnings.md
+depends_on:
+  - harness-verification
+
+</execution_context>
+
+<process>
+1. Try: invoke mcp__harness__run_skill with skill: "harness-execution"
+2. If MCP unavailable: follow the SKILL.md workflow provided above directly
+3. Pass through any arguments provided by the user
+</process>
+"""

package/dist/agents/{agents/skills/claude-code/harness-git-workflow/SKILL.md → commands/gemini-cli/harness/git-workflow.toml}

@@ -1,3 +1,17 @@
+# Generated by harness generate-slash-commands. Do not edit.
+description = "Git workflow best practices integrated with harness validation"
+prompt = """
+<context>
+Cognitive mode: meticulous-verifier
+Type: flexible
+</context>
+
+<objective>
+Git workflow best practices integrated with harness validation
+</objective>
+
+<execution_context>
+--- SKILL.md (agents/skills/claude-code/harness-git-workflow/SKILL.md) ---
 # Harness Git Workflow
 
 > Worktree setup, dependency installation, baseline verification, and branch finishing. Clean isolation for every workstream.
@@ -266,3 +280,46 @@ git branch -D feat/ws-experiment
 git worktree prune
 git worktree list    # ws-experiment no longer listed
 ```
+
+
+--- skill.yaml (agents/skills/claude-code/harness-git-workflow/skill.yaml) ---
+name: harness-git-workflow
+version: "1.0.0"
+description: Git workflow best practices integrated with harness validation
+cognitive_mode: meticulous-verifier
+triggers:
+  - manual
+  - on_pr
+  - on_commit
+platforms:
+  - claude-code
+  - gemini-cli
+tools:
+  - Bash
+  - Read
+  - Glob
+cli:
+  command: harness skill run harness-git-workflow
+  args:
+    - name: path
+      description: Project root path
+      required: false
+mcp:
+  tool: run_skill
+  input:
+    skill: harness-git-workflow
+    path: string
+type: flexible
+state:
+  persistent: false
+  files: []
+depends_on: []
+
+</execution_context>
+
+<process>
+1. Try: invoke mcp__harness__run_skill with skill: "harness-git-workflow"
+2. If MCP unavailable: follow the SKILL.md workflow provided above directly
+3. Pass through any arguments provided by the user
+</process>
+"""

package/dist/agents/{skills/gemini-cli/initialize-harness-project/SKILL.md → commands/gemini-cli/harness/initialize-project.toml}

@@ -1,3 +1,17 @@
+# Generated by harness generate-slash-commands. Do not edit.
+description = "Scaffold a new harness-compliant project"
+prompt = """
+<context>
+Cognitive mode: constructive-architect
+Type: flexible
+</context>
+
+<objective>
+Scaffold a new harness-compliant project
+</objective>
+
+<execution_context>
+--- SKILL.md (agents/skills/claude-code/initialize-harness-project/SKILL.md) ---
 # Initialize Harness Project
 
 > Scaffold a new harness-compliant project or migrate an existing project to the next adoption level. Assess current state, configure personas, generate AGENTS.md, and validate the result.
@@ -198,3 +212,46 @@ git commit -m "feat: migrate harness project to intermediate level with layers"
 - `.harness/learnings.md` capturing institutional knowledge
 - `harness validate` runs in CI pipeline
 - Custom skills for all common team workflows
+
+
+--- skill.yaml (agents/skills/claude-code/initialize-harness-project/skill.yaml) ---
+name: initialize-harness-project
+version: "1.0.0"
+description: Scaffold a new harness-compliant project
+cognitive_mode: constructive-architect
+triggers:
+  - manual
+  - on_project_init
+platforms:
+  - claude-code
+  - gemini-cli
+tools:
+  - Bash
+  - Read
+  - Write
+  - Glob
+cli:
+  command: harness skill run initialize-harness-project
+  args:
+    - name: path
+      description: Project root path
+      required: false
+mcp:
+  tool: run_skill
+  input:
+    skill: initialize-harness-project
+    path: string
+type: flexible
+state:
+  persistent: false
+  files: []
+depends_on: []
+
+</execution_context>
+
+<process>
+1. Try: invoke mcp__harness__run_skill with skill: "initialize-harness-project"
+2. If MCP unavailable: follow the SKILL.md workflow provided above directly
+3. Pass through any arguments provided by the user
+</process>
+"""

package/dist/agents/{skills/gemini-cli/harness-onboarding/SKILL.md → commands/gemini-cli/harness/onboarding.toml}

@@ -1,3 +1,17 @@
+# Generated by harness generate-slash-commands. Do not edit.
+description = "Onboard a new developer to a harness-managed project"
+prompt = """
+<context>
+Cognitive mode: advisory-guide
+Type: flexible
+</context>
+
+<objective>
+Onboard a new developer to a harness-managed project
+</objective>
+
+<execution_context>
+--- SKILL.md (agents/skills/claude-code/harness-onboarding/SKILL.md) ---
 # Harness Onboarding
 
 > Navigate an existing harness-managed project and generate a structured orientation for new team members. Map the codebase, understand constraints, identify the adoption level, and produce a summary that gets someone productive fast.
@@ -258,3 +272,45 @@ A mental model for where a team sits on the harness adoption curve. Not prescrip
 | 4     | **Self-improving** | Learnings accumulate in `.harness/learnings.md`. Agents reference past failures before planning. Institutional knowledge compounds across sessions and team members. |
 
 Most teams start at Level 1 and move up as they see the value. There is no pressure to reach Level 4 — each level delivers real benefits on its own.
+
+
+--- skill.yaml (agents/skills/claude-code/harness-onboarding/skill.yaml) ---
+name: harness-onboarding
+version: "1.0.0"
+description: Onboard a new developer to a harness-managed project
+cognitive_mode: advisory-guide
+triggers:
+  - manual
+  - on_project_init
+platforms:
+  - claude-code
+  - gemini-cli
+tools:
+  - Bash
+  - Read
+  - Glob
+cli:
+  command: harness skill run harness-onboarding
+  args:
+    - name: path
+      description: Project root path
+      required: false
+mcp:
+  tool: run_skill
+  input:
+    skill: harness-onboarding
+    path: string
+type: flexible
+state:
+  persistent: false
+  files: []
+depends_on: []
+
+</execution_context>
+
+<process>
+1. Try: invoke mcp__harness__run_skill with skill: "harness-onboarding"
+2. If MCP unavailable: follow the SKILL.md workflow provided above directly
+3. Pass through any arguments provided by the user
+</process>
+"""