create-hivemind-protocol 1.0.0

package/memory/agent-states/data.state.md

@@ -0,0 +1,19 @@
+# data — Agent State
+
+> Updated at the end of each session.
+
+## Last Session
+**Date**: —
+**Model used**: —
+
+## Current Focus
+—
+
+## Last Action
+—
+
+## Pending Items
+- [ ] Review project.json after CTO configures it
+
+## Notes for Next Session
+—

package/memory/agent-states/devops.state.md

@@ -0,0 +1,19 @@
+# devops — Agent State
+
+> Updated at the end of each session.
+
+## Last Session
+**Date**: —
+**Model used**: —
+
+## Current Focus
+—
+
+## Last Action
+—
+
+## Pending Items
+- [ ] Review project.json after CTO configures it
+
+## Notes for Next Session
+—

package/memory/agent-states/docs.state.md

@@ -0,0 +1,19 @@
+# docs — Agent State
+
+> Updated at the end of each session.
+
+## Last Session
+**Date**: —
+**Model used**: —
+
+## Current Focus
+—
+
+## Last Action
+—
+
+## Pending Items
+- [ ] Review project.json after CTO configures it
+
+## Notes for Next Session
+—

package/memory/agent-states/frontend-dev.state.md

@@ -0,0 +1,19 @@
+# frontend-dev — Agent State
+
+> Updated at the end of each session.
+
+## Last Session
+**Date**: —
+**Model used**: —
+
+## Current Focus
+—
+
+## Last Action
+—
+
+## Pending Items
+- [ ] Review project.json after CTO configures it
+
+## Notes for Next Session
+—

package/memory/agent-states/lead-dev.state.md

@@ -0,0 +1,19 @@
+# lead-dev — Agent State
+
+> Updated at the end of each session.
+
+## Last Session
+**Date**: —
+**Model used**: —
+
+## Current Focus
+—
+
+## Last Action
+—
+
+## Pending Items
+- [ ] Review project.json after CTO configures it
+
+## Notes for Next Session
+—

package/memory/agent-states/mobile.state.md

@@ -0,0 +1,19 @@
+# mobile — Agent State
+
+> Updated at the end of each session.
+
+## Last Session
+**Date**: —
+**Model used**: —
+
+## Current Focus
+—
+
+## Last Action
+—
+
+## Pending Items
+- [ ] Review project.json after CTO configures it
+
+## Notes for Next Session
+—

package/memory/agent-states/product-manager.state.md

@@ -0,0 +1,19 @@
+# product-manager — Agent State
+
+> Updated at the end of each session.
+
+## Last Session
+**Date**: —
+**Model used**: —
+
+## Current Focus
+—
+
+## Last Action
+—
+
+## Pending Items
+- [ ] Review project.json after CTO configures it
+
+## Notes for Next Session
+—

package/memory/agent-states/qa.state.md

@@ -0,0 +1,19 @@
+# qa — Agent State
+
+> Updated at the end of each session.
+
+## Last Session
+**Date**: —
+**Model used**: —
+
+## Current Focus
+—
+
+## Last Action
+—
+
+## Pending Items
+- [ ] Review project.json after CTO configures it
+
+## Notes for Next Session
+—

package/memory/agent-states/security.state.md

@@ -0,0 +1,19 @@
+# security — Agent State
+
+> Updated at the end of each session.
+
+## Last Session
+**Date**: —
+**Model used**: —
+
+## Current Focus
+—
+
+## Last Action
+—
+
+## Pending Items
+- [ ] Review project.json after CTO configures it
+
+## Notes for Next Session
+—

package/memory/blockers.md

@@ -0,0 +1,36 @@
+# Blockers
+
+> **Protocol**: Add blockers immediately when identified. Resolve by marking `[RESOLVED: YYYY-MM-DD]`.
+> Use `/blocker <description>` to add. Mark resolved when fixed.
+
+---
+
+## Active Blockers
+
+_None — project is freshly initialized._
+
+---
+
+## Resolved Blockers
+
+<!-- Move resolved blockers here with their resolution date and notes -->
+
+---
+
+## Format
+
+```
+### [YYYY-MM-DD] [SEVERITY] <Title>
+**Reported by**: <agent>
+**Blocking**: <what can't proceed>
+**Description**: <details>
+**Escalated to**: <agent if applicable>
+**Status**: [OPEN] / [IN-PROGRESS] / [RESOLVED: YYYY-MM-DD by <agent>]
+**Resolution**: <how it was resolved, if applicable>
+```
+
+### Severity Levels
+- `[CRITICAL]` — halts all progress on the affected system
+- `[HIGH]` — blocks a specific agent or feature
+- `[MEDIUM]` — workaround exists, but should be fixed soon
+- `[LOW]` — minor inconvenience, low urgency

package/memory/decisions.log

@@ -0,0 +1,14 @@
+# Decisions Log
+
+> **Protocol**: Append-only. Never edit or delete entries. Format strictly enforced.
+> **Format**: `[YYYY-MM-DD HH:MM] [AGENT] DECISION: <description> | REASON: <reason>`
+>
+> Use `/decision <agent> <text>` command to add entries.
+
+---
+
+## Log
+
+[YYYY-MM-DD HH:MM] [init] DECISION: HiveMind Protocol repository initialized | REASON: Project kickoff using HiveMind Protocol template
+
+<!-- Add new entries below this line -->

package/memory/handoff-queue.md

@@ -0,0 +1,17 @@
+# Handoff Queue
+
+> **Protocol**: Append-only. Mark completed items as `[DONE]`. Never delete entries.
+> Use `/handoff <from> <to> <task>` command to add entries.
+
+---
+
+## Queue
+
+### [YYYY-MM-DD] FROM: init → TO: cto
+**TASK**: Review project.json and configure for your project
+**CONTEXT**: HiveMind Protocol template initialized. All agents need project.json to be filled in (meta, stack, active agents) before starting work.
+**FILES**: `project.json`, `CLAUDE.md`
+**MODEL**: lite
+**STATUS**: [PENDING]
+
+<!-- Add new handoff entries below this line -->

package/memory/shared-context.md

@@ -0,0 +1,65 @@
+# Shared Context
+
+> **Protocol**: All agents must read this file at session start. Update with append only — never delete history. Use the format below.
+
+---
+
+## Current Project State
+
+**Last updated**: YYYY-MM-DD by [agent]
+**Phase**: Setup / Development / Staging / Production
+**Sprint**: 0
+
+### Active Focus
+_Describe what the team is currently working on in 1-3 sentences._
+
+### Recent Significant Changes
+_Brief summary of last 3 meaningful changes — agents should check CHANGELOG for full detail._
+
+---
+
+## Technical Snapshot
+
+### Stack
+_Fill in from `project.json > meta.stack` after project initialization._
+
+| Layer | Technology | Status |
+|-------|-----------|--------|
+| Frontend | — | — |
+| Backend | — | — |
+| Database | — | — |
+| Infra | — | — |
+| AI | — | — |
+
+### API Status
+| Endpoint group | Status | Owner |
+|----------------|--------|-------|
+| /auth | — | backend-dev |
+| /users | — | backend-dev |
+
+### Environment Status
+| Environment | Status | Last deployed |
+|------------|--------|--------------|
+| development | — | — |
+| staging | — | — |
+| production | — | — |
+
+---
+
+## Decisions in Effect
+
+_Key decisions that all agents should be aware of. Full log in `memory/decisions.log`._
+
+| Date | Decision | Made by |
+|------|----------|---------|
+| — | — | — |
+
+---
+
+## History
+
+_Append entries below when updating. Format: `[YYYY-MM-DD HH:MM] [agent] <summary of change>`_
+
+```
+[YYYY-MM-DD HH:MM] [init] Repository initialized from HiveMind Protocol template
+```

package/package.json

@@ -0,0 +1,40 @@
+{
+  "name": "create-hivemind-protocol",
+  "version": "1.0.0",
+  "description": "HiveMind Protocol — multi-agent context and behavior framework for AI-assisted software development",
+  "bin": {
+    "create-hivemind-protocol": "./bin/create.js"
+  },
+  "files": [
+    "bin",
+    "agents",
+    "assets",
+    "memory",
+    "tools",
+    "reports",
+    "CLAUDE.md",
+    "project.json",
+    "README.md",
+    "LICENSE"
+  ],
+  "keywords": [
+    "claude",
+    "ai-agents",
+    "llm",
+    "multi-agent",
+    "scaffolder",
+    "claude-code",
+    "agentic",
+    "template"
+  ],
+  "author": "Awi-24",
+  "license": "MIT",
+  "engines": {
+    "node": ">=16"
+  },
+  "repository": {
+    "type": "git",
+    "url": "https://github.com/Awi-24/HiveMind-Protocol.git"
+  },
+  "homepage": "https://github.com/Awi-24/HiveMind-Protocol#readme"
+}

package/project.json

@@ -0,0 +1,298 @@
+{
+  "$schema": "./agentos-schema.json",
+  "_comment": "HiveMind Protocol project definition file. Fill in all fields before starting. Boolean flags and arrays can be left empty.",
+
+  "meta": {
+    "name": "my-project",
+    "version": "0.1.0",
+    "description": "Short description of the project",
+    "repo": "https://github.com/org/repo",
+    "created": "YYYY-MM-DD",
+    "team": "team-name",
+    "stack": {
+      "language": ["typescript", "python"],
+      "framework": ["next.js", "fastapi"],
+      "database": ["postgresql", "redis"],
+      "cloud": ["aws"],
+      "infra": ["docker", "kubernetes"]
+    }
+  },
+
+  "agents": {
+    "active": [
+      "cto",
+      "lead-dev",
+      "backend-dev",
+      "frontend-dev",
+      "devops",
+      "security",
+      "qa"
+    ],
+    "available": [
+      "cto",
+      "lead-dev",
+      "product-manager",
+      "backend-dev",
+      "frontend-dev",
+      "devops",
+      "security",
+      "qa",
+      "data",
+      "docs",
+      "mobile",
+      "ai-ml"
+    ],
+    "profiles_dir": "./agents/"
+  },
+
+  "routing": {
+    "_comment": "Model routing by task complexity. Agents read this to select the correct model per task.",
+    "lite": {
+      "model": "claude-haiku-4-5-20251001",
+      "use_for": [
+        "reading files",
+        "writing reports",
+        "appending to logs",
+        "status checks",
+        "formatting markdown",
+        "updating agent states",
+        "handoff queue updates"
+      ],
+      "max_tokens": 4096
+    },
+    "standard": {
+      "model": "claude-sonnet-4-6",
+      "use_for": [
+        "writing code",
+        "debugging",
+        "implementing logic",
+        "API design",
+        "writing tests",
+        "code reviews",
+        "PR descriptions",
+        "CI/CD configuration"
+      ],
+      "max_tokens": 8192
+    },
+    "heavy": {
+      "model": "claude-opus-4-6",
+      "use_for": [
+        "architecture decisions",
+        "security audits",
+        "complex refactors spanning 3+ systems",
+        "cross-team design sessions",
+        "breaking change analysis",
+        "incident root cause analysis",
+        "multi-agent orchestration planning"
+      ],
+      "max_tokens": 16384
+    },
+    "default_tier": "standard",
+    "escalation_rule": "If a standard task takes more than 3 attempts, escalate to heavy and log in decisions.log"
+  },
+
+  "railguards": {
+    "max_loop_depth": 3,
+    "max_file_iterations_per_session": 5,
+    "require_memory_read_before_write": true,
+    "require_report_on_change": true,
+    "require_confirmation_before_destructive": true,
+    "forbidden_operations": [
+      "rm -rf",
+      "DROP TABLE",
+      "DROP DATABASE",
+      "TRUNCATE",
+      "git push --force",
+      "git reset --hard",
+      "kubectl delete namespace",
+      "terraform destroy"
+    ],
+    "forbidden_patterns_in_code": [
+      "eval(",
+      "exec(",
+      "shell=True",
+      "innerHTML =",
+      "dangerouslySetInnerHTML",
+      "password.*=.*['\"][^'\"]+['\"]"
+    ],
+    "on_loop_limit": "log to memory/blockers.md and escalate to CTO",
+    "on_forbidden_operation": "stop immediately, log to memory/blockers.md, notify user"
+  },
+
+  "communication": {
+    "_comment": "Token compression protocol. Inspired by https://github.com/JuliusBrussee/caveman — ~65% output reduction, ~46% input reduction via memory compression.",
+    "default_intensity": "full",
+    "intensity_levels": {
+      "lite": {
+        "token_reduction": "~40%",
+        "rules": "Keep articles; full sentences OK; drop filler words only",
+        "use_for": ["human-facing explanations", "user-visible output", "onboarding"]
+      },
+      "full": {
+        "token_reduction": "~60%",
+        "rules": "Drop articles; fragments allowed; no pleasantries; no preamble/postamble",
+        "use_for": ["standard agent-to-agent communication", "code reviews", "handoffs"]
+      },
+      "ultra": {
+        "token_reduction": "~75%",
+        "rules": "Abbreviations allowed (DB/auth/cfg/deps); arrows for causality (→); maximum compression",
+        "use_for": ["memory file writes", "log entries", "status checks", "internal agent chains"]
+      }
+    },
+    "auto_clarity_exception": {
+      "enabled": true,
+      "triggers": [
+        "security warnings severity HIGH or CRITICAL",
+        "irreversible operations: DROP, DELETE, rm -rf, force push",
+        "multi-step sequences with order-sensitive steps",
+        "escalations to the user"
+      ],
+      "marker_start": "[COMPRESSION SUSPENDED — CRITICAL]",
+      "marker_end": "[COMPRESSION RESUMED]"
+    },
+    "memory_compression": {
+      "enabled": true,
+      "target_reduction": "~46%",
+      "preserve": ["code blocks", "file paths", "URLs", "technical terms", "library names", "dates", "version numbers", "model IDs", "headings", "bullet structure"],
+      "drop": ["articles (a/an/the)", "filler (just/really/basically)", "hedging (might/could/perhaps)", "meta-phrases (It is worth noting that...)"]
+    },
+    "per_agent_overrides": {
+      "_comment": "Override default_intensity per agent slug. Leave empty to use default.",
+      "cto": "lite",
+      "product-manager": "lite",
+      "docs": "lite"
+    }
+  },
+
+  "commands": {
+    "_comment": "Custom slash commands available to all agents. See tools/custom-commands.md for details.",
+    "status": {
+      "syntax": "/status",
+      "description": "Summarize current state of all active agents",
+      "model_tier": "lite"
+    },
+    "handoff": {
+      "syntax": "/handoff <from-agent> <to-agent> <task>",
+      "description": "Initiate a formal agent handoff",
+      "model_tier": "lite"
+    },
+    "report": {
+      "syntax": "/report <agent> <summary>",
+      "description": "Log an entry to reports/CHANGELOG.md",
+      "model_tier": "lite"
+    },
+    "scaffold": {
+      "syntax": "/scaffold <template> [--name <project-name>] [--out <dir>]",
+      "description": "Bootstrap project structure from a template",
+      "model_tier": "standard",
+      "templates": ["nextjs", "fastapi", "react-native", "node-api", "monorepo"]
+    },
+    "blocker": {
+      "syntax": "/blocker <description> [--owner <agent>]",
+      "description": "Register a blocker in memory/blockers.md",
+      "model_tier": "lite"
+    },
+    "decision": {
+      "syntax": "/decision <agent> <text>",
+      "description": "Append a decision to memory/decisions.log",
+      "model_tier": "lite"
+    },
+    "audit": {
+      "syntax": "/audit [--agent <agent>] [--scope <files|deps|secrets>]",
+      "description": "Trigger a security audit by the Security agent",
+      "model_tier": "heavy"
+    },
+    "sprint": {
+      "syntax": "/sprint [--week <YYYY-Www>]",
+      "description": "Generate sprint report in reports/sprint-report.md",
+      "model_tier": "lite"
+    },
+    "focus": {
+      "syntax": "/focus <agent>",
+      "description": "Scope session to a specific agent — load profile, state, and context",
+      "model_tier": "lite"
+    },
+    "standup": {
+      "syntax": "/standup",
+      "description": "Generate daily standup summary across all active agents",
+      "model_tier": "lite"
+    },
+    "checkpoint": {
+      "syntax": "/checkpoint [--label <name>]",
+      "description": "Snapshot current state before a risky operation",
+      "model_tier": "lite"
+    },
+    "review": {
+      "syntax": "/review <file> [--agent <agent>] [--scope <logic|security|style|all>]",
+      "description": "Structured code review of a file by the appropriate agent",
+      "model_tier": "standard"
+    },
+    "hotfix": {
+      "syntax": "/hotfix <description> [--severity <critical|high>]",
+      "description": "Emergency fix workflow — fast-tracks through QA → Security → DevOps",
+      "model_tier": "standard"
+    },
+    "resolve": {
+      "syntax": "/resolve <blocker-title> [--resolution <text>]",
+      "description": "Close an active blocker in memory/blockers.md",
+      "model_tier": "lite"
+    },
+    "deploy": {
+      "syntax": "/deploy [--env <staging|production>] [--agent <agent>]",
+      "description": "Formalize QA → Security → DevOps deployment handoff chain",
+      "model_tier": "standard"
+    }
+  },
+
+  "mcps": {
+    "_comment": "MCP servers used in this project. See tools/mcp-catalog.md for setup instructions.",
+    "enabled": [],
+    "available": [
+      "github",
+      "filesystem",
+      "postgres",
+      "redis",
+      "notion",
+      "slack",
+      "jira",
+      "docker",
+      "aws",
+      "vercel",
+      "linear"
+    ]
+  },
+
+  "memory": {
+    "shared_context": "./memory/shared-context.md",
+    "decisions_log": "./memory/decisions.log",
+    "handoff_queue": "./memory/handoff-queue.md",
+    "blockers": "./memory/blockers.md",
+    "agent_states_dir": "./memory/agent-states/"
+  },
+
+  "reports": {
+    "changelog": "./reports/CHANGELOG.md",
+    "sprint_report": "./reports/sprint-report.md",
+    "audit_log": "./reports/audit-log.md"
+  },
+
+  "customization": {
+    "_comment": "Project-specific overrides for agent behavior",
+    "code_style": {
+      "indent": 2,
+      "quotes": "double",
+      "trailing_comma": true,
+      "line_length": 100
+    },
+    "git": {
+      "branch_pattern": "feat/<agent>/<short-description>",
+      "commit_pattern": "<type>(<agent>): <description>",
+      "require_pr_review": true
+    },
+    "environments": {
+      "development": { "auto_seed_db": true, "mock_external_services": true },
+      "staging": { "auto_seed_db": false, "mock_external_services": false },
+      "production": { "read_only_agents": ["qa", "docs", "data"] }
+    }
+  }
+}

package/reports/CHANGELOG.md

@@ -0,0 +1,32 @@
+# CHANGELOG
+
+> Collaborative changelog written by all agents. Append only — never delete entries.
+> Use `/report <agent> <summary>` to add entries.
+
+---
+
+## Entry Format
+
+```markdown
+## [YYYY-MM-DD] — <Agent>
+### <Type>: <Short title>
+- **What**: <what was done>
+- **Why**: <motivation>
+- **Impact**: <what changes for other agents>
+- **Files**: <list of files>
+- **Model used**: <lite|standard|heavy>
+```
+
+Types: `feat` | `fix` | `chore` | `docs` | `security` | `perf` | `refactor` | `test` | `infra`
+
+---
+
+## [YYYY-MM-DD] — init
+### chore: HiveMind Protocol repository initialized
+- **What**: Created HiveMind Protocol template structure with 12 agent profiles, shared memory system, tools catalog, and reports infrastructure
+- **Why**: Establish a multi-agent development framework for consistent AI-assisted development
+- **Impact**: All agents can now be initialized and begin working with shared memory and structured protocols
+- **Files**: All files in `agents/`, `memory/`, `tools/`, `reports/`, `CLAUDE.md`, `project.json`
+- **Model used**: standard
+
+<!-- Add new entries above this line, newest first -->