@kal-elsam/kairo-runtime 0.1.0

package/prompts/HARNESS_MINIMAL.md

@@ -0,0 +1,30 @@
+# Harness Installer — Minimal Mode
+
+Instala un harness mínimo para proyecto pequeño.
+
+Crea o actualiza:
+
+```txt
+AGENTS.md
+docs/ai/architecture.md
+docs/ai/testing.md
+docs/ai/git-workflow.md
+.cursor/rules/core.mdc
+.cursor/rules/testing.mdc
+```
+
+Reglas:
+
+- No instalar dependencias.
+- No duplicar reglas largas.
+- No crear `.cursor/agents` ni `docs/skills` salvo que el usuario lo pida.
+- Mantener `AGENTS.md` corto y accionable.
+- Documentar comandos reales detectados.
+- Bugs requieren regression test primero.
+- Features no triviales requieren mini spec.
+
+Flujo:
+
+```txt
+Requirement → Mini Spec → Test/Validation → Implementation → Review
+```

package/prompts/HARNESS_SPEC_SIZING_UPGRADE.md

@@ -0,0 +1,41 @@
+# Harness Spec Sizing Upgrade
+
+Use this prompt to add spec complexity classification to an existing harness.
+
+## Add or update
+
+```txt
+docs/ai/spec-sizing.md
+docs/ai/spec-intake.md
+docs/ai/spec-escalation.md
+docs/specs/templates/basic-spec.md
+docs/specs/templates/standard-spec.md
+docs/specs/templates/complex-spec.md
+docs/skills/spec-complexity-classifier.md
+docs/skills/spec-intake.md
+docs/skills/spec-escalation-review.md
+```
+
+## Update adapters
+
+Add spec sizing support to:
+
+```txt
+.cursor/commands/spec-size.md
+.cursor/commands/spec-intake.md
+.cursor/commands/spec-escalate.md
+.opencode/commands/spec-size.md
+.opencode/commands/spec-intake.md
+.opencode/commands/spec-escalate.md
+.codex/skills/spec-complexity-classifier/SKILL.md
+.claude/skills/spec-complexity-classifier/SKILL.md
+.pi/prompts/spec-size.md
+```
+
+## Rule
+
+```txt
+Start with the smallest safe spec.
+Escalate when risk or ambiguity increases.
+Never use a basic spec for critical work.
+```

package/prompts/HARNESS_STANDARD.md

@@ -0,0 +1,35 @@
+# Harness Installer — Standard Mode
+
+Instala un harness estándar para app real.
+
+Crea:
+
+```txt
+AGENTS.md
+docs/ai/
+docs/skills/
+.cursor/rules/
+.cursor/commands/
+```
+
+Debe incluir:
+
+- `docs/ai/architecture.md`
+- `docs/ai/conventions.md`
+- `docs/ai/testing.md`
+- `docs/ai/git-workflow.md`
+- `docs/ai/harness.md`
+- `docs/ai/spec-driven-development.md`
+- `docs/ai/test-driven-development.md`
+- `docs/ai/agent-workflow.md`
+- `docs/ai/decision-log.md`
+- `docs/ai/ui.md` si hay UI
+- `docs/ai/api.md` si hay API
+
+Flujo obligatorio:
+
+```txt
+Requirement → Spec → Plan → Tests failing first → Implementation → Validation → Review → Human approval
+```
+
+No crear agentes especializados salvo que el proyecto lo justifique.

package/prompts/HARNESS_UNIVERSAL_PARITY.md

@@ -0,0 +1,56 @@
+# Harness Installer — Universal Adapter Parity
+
+Use this prompt when the project must support multiple agents/tools without lock-in.
+
+## Goal
+
+Ensure the project uses a universal core and adapter parity.
+
+## Core
+
+```txt
+AGENTS.md
+docs/ai/
+docs/skills/
+docs/specs/
+evals/
+```
+
+## Adapters
+
+```txt
+.opencode/
+.cursor/
+.codex/
+.claude/
+.github/
+.gemini/
+.pi/
+.gentle-ai/
+```
+
+## Rules
+
+```txt
+No adapter is primary by authority.
+An adapter can be primary only by workflow preference.
+The core universal remains governance-primary.
+```
+
+## Required output
+
+Create or update:
+
+```txt
+docs/ai/adapter-parity.md
+docs/ai/governance.md
+docs/skills/adapter-parity-review.md
+```
+
+Then verify:
+
+- OpenCode supports SDD/TDD/evals/loops/checkpoint.
+- Cursor supports SDD/TDD/evals/loops/checkpoint.
+- Codex supports SDD/TDD/evals/loops/checkpoint.
+- Claude supports SDD/TDD/evals/loops/checkpoint.
+- Pi/Gemini/Copilot have at least pointer or prompt support.

package/repo-template/.claude/agents/adapter-parity-reviewer.md

@@ -0,0 +1,17 @@
+# Adapter Parity Reviewer
+
+## Role
+
+Review whether Claude adapter stays aligned with the universal harness.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/adapter-parity.md`
+- `docs/ai/governance.md`
+
+## Rules
+
+- Claude is reasoning-strong, not governance-primary.
+- Do not duplicate long core rules here.
+- Report gaps and contradictions.

package/repo-template/.claude/agents/architect.md

@@ -0,0 +1,17 @@
+# architect
+
+## Role
+
+Review architecture using `docs/ai/architecture.md`.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/`
+- `docs/skills/`
+
+## Rules
+
+- Respect context budget.
+- Do not contradict the core harness.
+- Report risks and validation.

package/repo-template/.claude/agents/eval-engineer.md

@@ -0,0 +1,17 @@
+# eval-engineer
+
+## Role
+
+Design AI evals using `docs/ai/evals.md`.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/`
+- `docs/skills/`
+
+## Rules
+
+- Respect context budget.
+- Do not contradict the core harness.
+- Report risks and validation.

package/repo-template/.claude/agents/reviewer.md

@@ -0,0 +1,17 @@
+# reviewer
+
+## Role
+
+Perform final review using `docs/ai/` and `docs/skills/`.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/`
+- `docs/skills/`
+
+## Rules
+
+- Respect context budget.
+- Do not contradict the core harness.
+- Report risks and validation.

package/repo-template/.claude/agents/sdd-reviewer.md

@@ -0,0 +1,17 @@
+# sdd-reviewer
+
+## Role
+
+Review specs for contradictions, missing acceptance criteria and scope creep.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/`
+- `docs/skills/`
+
+## Rules
+
+- Respect context budget.
+- Do not contradict the core harness.
+- Report risks and validation.

package/repo-template/.claude/agents/sdd-spec-writer.md

@@ -0,0 +1,17 @@
+# sdd-spec-writer
+
+## Role
+
+Write or update specs based on `docs/ai/spec-driven-development.md`.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/`
+- `docs/skills/`
+
+## Rules
+
+- Respect context budget.
+- Do not contradict the core harness.
+- Report risks and validation.

package/repo-template/.claude/skills/checkpoint/SKILL.md

@@ -0,0 +1,13 @@
+# CHECKPOINT Skill
+
+Use before finalizing. Read `docs/ai/git-workflow.md`.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/`
+- `docs/skills/`
+
+## Rule
+
+This skill is an adapter. The core documentation wins.

package/repo-template/.claude/skills/eval-design/SKILL.md

@@ -0,0 +1,3 @@
+# eval-design
+
+Read and follow `docs/skills/eval-design.md`. The universal core governs this adapter skill.

package/repo-template/.claude/skills/evals/SKILL.md

@@ -0,0 +1,13 @@
+# EVALS Skill
+
+Use for AI evals. Read `docs/ai/evals.md`.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/`
+- `docs/skills/`
+
+## Rule
+
+This skill is an adapter. The core documentation wins.

package/repo-template/.claude/skills/loop-engineering/SKILL.md

@@ -0,0 +1,16 @@
+# Loop Engineering Skill
+
+Use for bounded loop design, review and debugging.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/loops.md`
+- `docs/ai/loop-policy.md`
+- `docs/ai/loop-observability.md`
+
+## Rules
+
+- No autonomous loop without boundaries.
+- Escalate security/auth/data/infra.
+- Report iterations, failures, repairs and risks.

package/repo-template/.claude/skills/maintainability-review/SKILL.md

@@ -0,0 +1,3 @@
+# maintainability-review
+
+Read and follow `docs/skills/maintainability-review.md`. The universal core governs this adapter skill.

package/repo-template/.claude/skills/quality-gate-review/SKILL.md

@@ -0,0 +1,3 @@
+# quality-gate-review
+
+Read and follow `docs/skills/quality-gate-review.md`. The universal core governs this adapter skill.

package/repo-template/.claude/skills/sdd/SKILL.md

@@ -0,0 +1,13 @@
+# SDD Skill
+
+Use for spec-driven development. Read `docs/ai/spec-driven-development.md`.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/`
+- `docs/skills/`
+
+## Rule
+
+This skill is an adapter. The core documentation wins.

package/repo-template/.claude/skills/spec-complexity-classifier/SKILL.md

@@ -0,0 +1,10 @@
+# spec-complexity-classifier
+
+Read and follow:
+
+```txt
+docs/skills/spec-complexity-classifier.md
+```
+
+Use the universal core as source of truth.
+Do not implement before classifying the spec level.

package/repo-template/.claude/skills/spec-escalation-review/SKILL.md

@@ -0,0 +1,10 @@
+# spec-escalation-review
+
+Read and follow:
+
+```txt
+docs/skills/spec-escalation-review.md
+```
+
+Use the universal core as source of truth.
+Do not implement before classifying the spec level.

package/repo-template/.claude/skills/spec-intake/SKILL.md

@@ -0,0 +1,10 @@
+# spec-intake
+
+Read and follow:
+
+```txt
+docs/skills/spec-intake.md
+```
+
+Use the universal core as source of truth.
+Do not implement before classifying the spec level.

package/repo-template/.claude/skills/tdd/SKILL.md

@@ -0,0 +1,13 @@
+# TDD Skill
+
+Use for test-driven development. Read `docs/ai/test-driven-development.md`.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/`
+- `docs/skills/`
+
+## Rule
+
+This skill is an adapter. The core documentation wins.

package/repo-template/.claude/skills/trust-review/SKILL.md

@@ -0,0 +1,3 @@
+# trust-review
+
+Read and follow `docs/skills/trust-review.md`. The universal core governs this adapter skill.

package/repo-template/.codex/skills/adapter-parity/SKILL.md

@@ -0,0 +1,29 @@
+# Adapter Parity Skill
+
+Use this skill to verify that Codex is aligned with the universal harness.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/adapter-parity.md`
+- `docs/ai/tool-adapters.md`
+- `docs/ai/governance.md`
+
+## Workflow
+
+1. Read `AGENTS.md`.
+2. Read `docs/ai/adapter-parity.md`.
+3. Check Codex skills under `.codex/skills/`.
+4. Confirm SDD/TDD/evals/checkpoint support.
+5. Confirm no duplicated long instructions.
+6. Report gaps.
+
+## Output
+
+```txt
+Codex adapter status:
+Capabilities:
+Gaps:
+Conflicts:
+Suggested changes:
+```

package/repo-template/.codex/skills/checkpoint/SKILL.md

@@ -0,0 +1,36 @@
+# Checkpoint Skill
+
+Use this skill before finalizing work.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/git-workflow.md`
+- `docs/ai/harness.md`
+
+## Workflow
+
+1. Review diff.
+2. Check secrets and accidental files.
+3. Run lint.
+4. Run format.
+5. Run typecheck.
+6. Run tests.
+7. Run evals if applicable.
+8. Run build.
+9. Propose conventional commit.
+10. Wait for human approval.
+
+## Output
+
+```txt
+Diff:
+Lint:
+Format:
+Types:
+Tests:
+Evals:
+Build:
+Commit:
+Risks:
+```

package/repo-template/.codex/skills/eval-design/SKILL.md

@@ -0,0 +1,3 @@
+# eval-design
+
+Read and follow `docs/skills/eval-design.md`. The universal core governs this adapter skill.

package/repo-template/.codex/skills/evals/SKILL.md

@@ -0,0 +1,28 @@
+# AI Evals Skill
+
+Use this skill for prompts, agents, LLM workflows, RAG and tool calling.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/evals.md`
+- `evals/`
+
+## Workflow
+
+1. Identify AI behavior.
+2. Create golden cases or regression cases.
+3. Validate schema adherence.
+4. Validate tool-call correctness.
+5. Validate safety constraints.
+6. Report cost/latency if applicable.
+
+## Output
+
+```txt
+Eval type:
+Cases added:
+Expected behavior:
+Validation:
+Risks:
+```

package/repo-template/.codex/skills/loop-engineering/SKILL.md

@@ -0,0 +1,24 @@
+# Loop Engineering Skill
+
+Use this skill for bounded repair, validation and learning loops in Codex.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/loops.md`
+- `docs/ai/loop-policy.md`
+- `docs/ai/loop-observability.md`
+
+## Workflow
+
+1. Identify loop type.
+2. Define input and max attempts.
+3. Execute one bounded iteration.
+4. Validate with tests/evals.
+5. Stop if max attempts or no progress.
+6. Escalate high-impact issues.
+7. Report final status.
+
+## Rule
+
+No autonomous loop without boundaries.

package/repo-template/.codex/skills/maintainability-review/SKILL.md

@@ -0,0 +1,3 @@
+# maintainability-review
+
+Read and follow `docs/skills/maintainability-review.md`. The universal core governs this adapter skill.

package/repo-template/.codex/skills/quality-gate-review/SKILL.md

@@ -0,0 +1,3 @@
+# quality-gate-review
+
+Read and follow `docs/skills/quality-gate-review.md`. The universal core governs this adapter skill.

package/repo-template/.codex/skills/sdd/SKILL.md

@@ -0,0 +1,30 @@
+# SDD Skill
+
+Use this skill for Spec-Driven Development.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/spec-driven-development.md`
+- `docs/ai/harness.md`
+- `docs/specs/`
+
+## Workflow
+
+1. Read `AGENTS.md`.
+2. Read `docs/ai/spec-driven-development.md`.
+3. Create or update the spec in `docs/specs/`.
+4. Define acceptance criteria.
+5. Create technical plan.
+6. Break work into implementation tasks.
+7. Do not implement until scope is clear.
+
+## Output
+
+```txt
+Spec path:
+Acceptance criteria:
+Technical plan:
+Tasks:
+Risks:
+```

package/repo-template/.codex/skills/spec-complexity-classifier/SKILL.md

@@ -0,0 +1,10 @@
+# spec-complexity-classifier
+
+Read and follow:
+
+```txt
+docs/skills/spec-complexity-classifier.md
+```
+
+Use the universal core as source of truth.
+Do not implement before classifying the spec level.

package/repo-template/.codex/skills/spec-escalation-review/SKILL.md

@@ -0,0 +1,10 @@
+# spec-escalation-review
+
+Read and follow:
+
+```txt
+docs/skills/spec-escalation-review.md
+```
+
+Use the universal core as source of truth.
+Do not implement before classifying the spec level.

package/repo-template/.codex/skills/spec-intake/SKILL.md

@@ -0,0 +1,10 @@
+# spec-intake
+
+Read and follow:
+
+```txt
+docs/skills/spec-intake.md
+```
+
+Use the universal core as source of truth.
+Do not implement before classifying the spec level.

package/repo-template/.codex/skills/tdd/SKILL.md

@@ -0,0 +1,27 @@
+# TDD Skill
+
+Use this skill for behavior-changing code.
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/test-driven-development.md`
+- `docs/ai/testing.md`
+
+## Workflow
+
+1. Identify expected behavior.
+2. Write failing test first.
+3. Confirm it fails for the right reason.
+4. Implement minimal change.
+5. Run related tests.
+6. Run full validation if needed.
+
+## Output
+
+```txt
+Failing test:
+Implementation:
+Validation:
+Coverage impact:
+```

package/repo-template/.codex/skills/trust-review/SKILL.md

@@ -0,0 +1,3 @@
+# trust-review
+
+Read and follow `docs/skills/trust-review.md`. The universal core governs this adapter skill.

package/repo-template/.cursor/agents/adapter-parity-reviewer.md

@@ -0,0 +1,29 @@
+---
+name: Adapter Parity Reviewer
+description: Reviews adapter support across OpenCode, Cursor, Codex, Claude, Gemini, Pi and Copilot.
+---
+
+# Adapter Parity Reviewer
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/adapter-parity.md`
+- `docs/ai/tool-adapters.md`
+- `docs/ai/governance.md`
+
+## Rules
+
+- Do not modify application code.
+- Do not promote any adapter above the core.
+- Identify missing capabilities.
+- Recommend minimal adapter changes.
+
+## Output
+
+```txt
+Adapters reviewed:
+Parity gaps:
+Conflicts:
+Recommended changes:
+```