@vectorize-io/self-driving-agents 0.0.24 → 0.0.26

package/README.md

@@ -1,136 +1,527 @@
-# Self-Driving Agents
+<p align="center">
+  <img src="public/sda-banner.jpg" alt="Self-Driving Agents" />
+</p>
 
-Agents that learn from every conversation and get better over time.
+<p align="center">
+  <a href="https://www.npmjs.com/package/@vectorize-io/self-driving-agents"><img src="https://img.shields.io/npm/v/@vectorize-io/self-driving-agents?style=flat-square" alt="npm version" /></a>
+  <a href="https://www.npmjs.com/package/@vectorize-io/self-driving-agents"><img src="https://img.shields.io/npm/dm/@vectorize-io/self-driving-agents?style=flat-square" alt="npm downloads" /></a>
+  <a href="https://github.com/vectorize-io/self-driving-agents/stargazers"><img src="https://img.shields.io/github/stars/vectorize-io/self-driving-agents?style=flat-square" alt="GitHub stars" /></a>
+  <a href="https://github.com/vectorize-io/self-driving-agents/commits"><img src="https://img.shields.io/github/last-commit/vectorize-io/self-driving-agents?style=flat-square" alt="Last commit" /></a>
+  <a href="https://github.com/vectorize-io/self-driving-agents/issues"><img src="https://img.shields.io/github/issues/vectorize-io/self-driving-agents?style=flat-square" alt="Open issues" /></a>
+  <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-yellow.svg?style=flat-square" alt="License: MIT" /></a>
+  <a href="https://vectorize-io.github.io/self-driving-agents/"><img src="https://img.shields.io/badge/docs-online-blue?style=flat-square" alt="Docs" /></a>
+</p>
 
-No retraining, no manual updates — just use them and they improve. Powered by [Hindsight](https://github.com/vectorize-io/hindsight), so agent memory is portable across harnesses.
+## A Complete AI Workforce in a Box
 
-📖 **Docs:** <https://vectorize-io.github.io/self-driving-agents/>
+179 ready-to-use, self-learning AI agents organized into 13 departments. Powered by [Hindsight](https://github.com/vectorize-io/hindsight).
 
-## Quick start
-
-### Claude Chat / Cowork
+**Divisions:** [Design](#design) · [Engineering](#engineering) · [Finance](#finance) · [Game Development](#game-development) · [Marketing](#marketing) · [Paid Media](#paid-media) · [Product](#product) · [Project Management](#project-management) · [Sales](#sales) · [Spatial Computing](#spatial-computing) · [Specialized](#specialized) · [Support](#support) · [Testing](#testing)
 
 ```bash
+# Built-in agent
 npx @vectorize-io/self-driving-agents install marketing/seo --harness claude
+
+# Local directory or any GitHub repo
+npx @vectorize-io/self-driving-agents install ./my-agent --harness claude
+npx @vectorize-io/self-driving-agents install my-org/my-repo/path --harness claude
+
+# Start blank, build through conversation
+npx @vectorize-io/self-driving-agents install my-agent --harness claude --empty
 ```
 
-The CLI prompts for your Hindsight connection (Cloud or self-hosted), creates a skill zip, and tells you how to upload it. Then type `/<agent-name>` in any conversation to activate the agent.
+Harnesses: `claude` · `claude-code` · `openclaw` · `nemoclaw` · `hermes`
 
-### OpenClaw
+**Type `/<agent-name>` in any conversation to activate.** Claude Chat / Cowork only: upload the generated skill via `Customize → Skills → Upload`, then allowlist the API host at `Settings → Capabilities`.
 
-```bash
-npx @vectorize-io/self-driving-agents install marketing/seo --harness openclaw
-```
+---
 
-### NemoClaw
+## Roster
 
-```bash
-npx @vectorize-io/self-driving-agents install marketing/seo --harness nemoclaw
-```
+### Design
 
-That's it. The CLI fetches the agent from this repo, sets up its memory bank, ingests the seed knowledge, and registers it with your harness. Follow the on-screen instructions to start chatting.
+#### [`design/ux`](design/ux/) — UX research, UI, architecture
 
-You can also point at a local directory or any GitHub repo:
+| Agent | Vibe |
+|-------|------|
+| 🎨 [UI Designer](design/ux/ui-designer.md) | Beautiful, consistent, accessible interfaces. |
+| 📐 [UX Architect](design/ux/ux-architect.md) | Developer-friendly CSS foundations and implementation paths. |
+| 🔬 [UX Researcher](design/ux/ux-researcher.md) | Real user data, not assumptions. |
 
-```bash
-npx @vectorize-io/self-driving-agents install ./my-agent --harness claude
-npx @vectorize-io/self-driving-agents install my-org/my-repo/my-agent --harness openclaw
-```
+#### [`design/visual`](design/visual/) — Brand, imagery, motion, delight
 
-Or start with a blank agent — no template, no seed content — using `--empty`:
+| Agent | Vibe |
+|-------|------|
+| 🎨 [Brand Guardian](design/visual/brand-guardian.md) | Fiercest protector of your brand. |
+| 📷 [Image Prompt Engineer](design/visual/image-prompt-engineer.md) | Precise prompts for stunning AI imagery. |
+| 🌈 [Inclusive Visuals Specialist](design/visual/inclusive-visuals-specialist.md) | Defeats AI bias for culturally accurate imagery. |
+| 🎬 [Visual Storyteller](design/visual/visual-storyteller.md) | Visual narratives that move people. |
+| ✨ [Whimsy Injector](design/visual/whimsy-injector.md) | The unexpected moments of delight. |
 
-```bash
-npx @vectorize-io/self-driving-agents install my-agent --harness claude-code --empty
-```
+### Engineering
 
-The first positional becomes the agent name. The CLI provisions the bank, sets up the harness, and lets you build everything up from the conversation.
+#### [`engineering/ai`](engineering/ai/) — AI/ML
 
-## How it works
+| Agent | Vibe |
+|-------|------|
+| 🤖 [AI Engineer](engineering/ai/ai-engineer.md) | Turns ML models into production features that actually scale. |
+| ⚡ [Autonomous Optimization Architect](engineering/ai/autonomous-optimization-architect.md) | System governor that makes things faster without bankrupting you. |
+| 📧 [Email Intelligence Engineer](engineering/ai/email-intelligence-engineer.md) | Turns messy MIME into reasoning-ready context. |
+| 🔧 [Filament Optimization Specialist](engineering/ai/filament-optimization-specialist.md) | Streamlines complex Filament admin environments. |
+| 🎙️ [Voice AI Integration Engineer](engineering/ai/voice-ai-integration-engineer.md) | Turns raw audio into structured, production-ready text. |
 
-1. You chat with the agent
-2. Conversations are automatically retained into memory
-3. The agent builds knowledge pages that update themselves after each conversation
-4. Next session, the agent reads its updated pages — it remembers what works and what you prefer
+#### [`engineering/architecture`](engineering/architecture/) — System & code architecture
 
-You never tell it to "save" or "remember." It decides what matters and keeps itself current.
+| Agent | Vibe |
+|-------|------|
+| ⚡ [Rapid Prototyper](engineering/architecture/rapid-prototyper.md) | Working prototype before the meeting's over. |
+| 💎 [Senior Developer](engineering/architecture/senior-developer.md) | Laravel, Livewire, Three.js, advanced CSS. |
+| 🏛️ [Software Architect](engineering/architecture/software-architect.md) | Systems that survive the team that built them. |
 
-## Available agents
+#### [`engineering/backend`](engineering/backend/) — APIs, data, databases
 
-### [marketing/](marketing/)
+| Agent | Vibe |
+|-------|------|
+| 🧬 [AI Data Remediation Engineer](engineering/backend/ai-data-remediation-engineer.md) | Fixes broken data with surgical AI precision. |
+| 🏗️ [Backend Architect](engineering/backend/backend-architect.md) | Databases, APIs, cloud, scale. |
+| 🧱 [CMS Developer](engineering/backend/cms-developer.md) | Drupal & WordPress at enterprise scale. |
+| 🔧 [Data Engineer](engineering/backend/data-engineer.md) | Pipelines that turn raw data into analytics-ready assets. |
+| 🗄️ [Database Optimizer](engineering/backend/database-optimizer.md) | Indexes, query plans, schema design — no 3am pages. |
 
-Full marketing team — install the whole department or pick a specialty.
+#### [`engineering/frontend`](engineering/frontend/) — Web & mobile UI
 
-| Install | What you get |
-|---------|-------------|
-| `marketing` | Generalist — all 30 agent knowledge files across every specialty |
-| `marketing/seo` | SEO specialist, search optimizer, citation strategist |
-| `marketing/social-media` | Platform strategists for TikTok, Instagram, LinkedIn, X, Reddit |
-| `marketing/content` | Content creator, growth hacker, book co-author |
-| `marketing/ecommerce` | Cross-border commerce, livestream selling |
-| `marketing/china-market` | WeChat, Douyin, Xiaohongshu, Baidu, Bilibili |
+| Agent | Vibe |
+|-------|------|
+| 🖥️ [Frontend Developer](engineering/frontend/frontend-developer.md) | Responsive, accessible web apps, pixel-perfect. |
+| 📲 [Mobile App Builder](engineering/frontend/mobile-app-builder.md) | Native-quality iOS and Android, fast. |
+| 💬 [WeChat Mini Program Developer](engineering/frontend/wechat-mini-program-developer.md) | Mini Programs that thrive in the WeChat ecosystem. |
 
-## Create your own
+#### [`engineering/ops`](engineering/ops/) — DevOps, SRE, firmware, incidents
 
-An agent is just a directory:
+| Agent | Vibe |
+|-------|------|
+| ⚙️ [DevOps Automator](engineering/ops/devops-automator.md) | Infra automation so the team ships faster and sleeps better. |
+| 🔩 [Embedded Firmware Engineer](engineering/ops/embedded-firmware-engineer.md) | Production-grade firmware for hardware that can't crash. |
+| 🚨 [Incident Response Commander](engineering/ops/incident-response-commander.md) | Production chaos → structured resolution. |
+| 🛡️ [SRE](engineering/ops/sre.md) | Error budgets fund velocity — spend them wisely. |
 
-```
-my-agent/
-  bank-template.json     # optional — configures memory bank and knowledge pages
-  seo-playbook.md        # any .md/.txt files become seed knowledge
-  advanced-tips.md
-```
+#### [`engineering/security`](engineering/security/) — AppSec, smart contracts, detection
 
-Nest directories for multi-level agents:
+| Agent | Vibe |
+|-------|------|
+| 🔒 [Security Engineer](engineering/security/security-engineer.md) | Threat modeling, code review, hunting — holds under adversaries. |
+| ⛓️ [Solidity Smart Contract Engineer](engineering/security/solidity-smart-contract-engineer.md) | Battle-hardened EVM developer. |
+| 🎯 [Threat Detection Engineer](engineering/security/threat-detection-engineer.md) | Catches attackers after they bypass prevention. |
 
-```
-my-team/
-  bank-template.json     # install my-team → everything below
-  specialist-a/
-    bank-template.json   # install my-team/specialist-a → just this
-    reference.md
-  specialist-b/
-    bank-template.json
-    guide.md
-```
+#### [`engineering/workflow`](engineering/workflow/) — Code review, Git, docs, onboarding
+
+| Agent | Vibe |
+|-------|------|
+| 👁️ [Code Reviewer](engineering/workflow/code-reviewer.md) | Mentor, not gatekeeper — every comment teaches. |
+| 🧭 [Codebase Onboarding Engineer](engineering/workflow/codebase-onboarding-engineer.md) | New devs productive — read code, trace paths, state facts. |
+| 🔗 [Feishu Integration Developer](engineering/workflow/feishu-integration-developer.md) | Feishu/Lark integrations — bots, approvals, data sync, SSO. |
+| 🌿 [Git Workflow Master](engineering/workflow/git-workflow-master.md) | Clean history, atomic commits, branches that tell a story. |
+| 🪡 [Minimal Change Engineer](engineering/workflow/minimal-change-engineer.md) | Smallest diff that solves the problem. |
+| 📚 [Technical Writer](engineering/workflow/technical-writer.md) | Docs developers actually read. |
+
+### Finance
+
+#### [`finance/accounting`](finance/accounting/) — Bookkeeping & tax
+
+| Agent | Vibe |
+|-------|------|
+| 📒 [Bookkeeper & Controller](finance/accounting/bookkeeper-controller.md) | Every penny accounted, every close on time. |
+| 🏛️ [Tax Strategist](finance/accounting/tax-strategist.md) | Every legal dollar of savings. |
+
+#### [`finance/analysis`](finance/analysis/) — FP&A, financial analysis, investment research
+
+| Agent | Vibe |
+|-------|------|
+| 📊 [Financial Analyst](finance/analysis/financial-analyst.md) | Spreadsheets → strategy. |
+| 📈 [FP&A Analyst](finance/analysis/fpa-analyst.md) | The budget whisperer. |
+| 🔍 [Investment Researcher](finance/analysis/investment-researcher.md) | Alpha in the footnotes, risk in the narratives. |
+
+### Game Development
+
+#### [`game-development/blender`](game-development/blender/)
+
+| Agent | Vibe |
+|-------|------|
+| 🧩 [Blender Add-on Engineer](game-development/blender/addon-engineer.md) | Pipeline drudgery → one-click tools. |
+
+#### [`game-development/core`](game-development/core/) — Engine-agnostic
+
+| Agent | Vibe |
+|-------|------|
+| 🎵 [Game Audio Engineer](game-development/core/game-audio-engineer.md) | Every sound alive in the world. |
+| 🎮 [Game Designer](game-development/core/game-designer.md) | Loops, levers, motivations. |
+| 🗺️ [Level Designer](game-development/core/level-designer.md) | Authored space tells the story. |
+| 📖 [Narrative Designer](game-development/core/narrative-designer.md) | Narrative and gameplay inseparable. |
+| 🎨 [Technical Artist](game-development/core/technical-artist.md) | Art vision ↔ engine reality. |
+
+#### [`game-development/godot`](game-development/godot/)
+
+| Agent | Vibe |
+|-------|------|
+| 🎯 [Godot Gameplay Scripter](game-development/godot/gameplay-scripter.md) | Godot 4 with software-architect discipline. |
+| 🌐 [Godot Multiplayer Engineer](game-development/godot/multiplayer-engineer.md) | MultiplayerAPI made seamless. |
+| 💎 [Godot Shader Developer](game-development/godot/shader-developer.md) | Light and pixels through Godot shading. |
+
+#### [`game-development/roblox-studio`](game-development/roblox-studio/)
+
+| Agent | Vibe |
+|-------|------|
+| 👤 [Roblox Avatar Creator](game-development/roblox-studio/roblox-avatar-creator.md) | UGC pipeline from rig to Marketplace. |
+| 🎪 [Roblox Experience Designer](game-development/roblox-studio/roblox-experience-designer.md) | Engagement loops and monetization. |
+| 🔧 [Roblox Systems Scripter](game-development/roblox-studio/roblox-systems-scripter.md) | Luau, client-server security, scale. |
+
+#### [`game-development/unity`](game-development/unity/)
+
+| Agent | Vibe |
+|-------|------|
+| 🏛️ [Unity Architect](game-development/unity/architect.md) | Data-driven, decoupled, no spaghetti. |
+| 🛠️ [Unity Editor Tool Developer](game-development/unity/editor-tool-developer.md) | Custom editor tools that save hours. |
+| 🔗 [Unity Multiplayer Engineer](game-development/unity/multiplayer-engineer.md) | Netcode that feels local. |
+| ✨ [Unity Shader Graph Artist](game-development/unity/shader-graph-artist.md) | Shader Graph + custom render passes. |
+
+#### [`game-development/unreal-engine`](game-development/unreal-engine/)
+
+| Agent | Vibe |
+|-------|------|
+| 🌐 [Unreal Multiplayer Architect](game-development/unreal-engine/unreal-multiplayer-architect.md) | Server-authoritative, feels lag-free. |
+| ⚙️ [Unreal Systems Engineer](game-development/unreal-engine/unreal-systems-engineer.md) | C++/Blueprint for AAA Unreal. |
+| 🎨 [Unreal Technical Artist](game-development/unreal-engine/unreal-technical-artist.md) | Niagara, Material Editor, PCG. |
+| 🌍 [Unreal World Builder](game-development/unreal-engine/unreal-world-builder.md) | World Partition, Nanite, procedural foliage. |
+
+### Marketing
+
+#### [`marketing/china-market`](marketing/china-market/) — Douyin, WeChat, Xiaohongshu, Bilibili, Weibo, Zhihu, Kuaishou
+
+| Agent | Vibe |
+|-------|------|
+| 🎬 [Bilibili Content Strategist](marketing/china-market/bilibili-content-strategist.md) | Fluent danmaku, grows your brand on B站. |
+| 🛒 [China E-Commerce Operator](marketing/china-market/china-ecommerce-operator.md) | Taobao, Tmall, Pinduoduo, JD — like a native operator. |
+| 🇨🇳 [China Market Localization Strategist](marketing/china-market/china-market-localization-strategist.md) | China trend chaos → precision marketing. |
+| 🎵 [Douyin Strategist](marketing/china-market/douyin-strategist.md) | Masters the Douyin algorithm. |
+| 🎥 [Kuaishou Strategist](marketing/china-market/kuaishou-strategist.md) | Grassroots audiences and live commerce on 快手. |
+| 🔒 [Private Domain Operator](marketing/china-market/private-domain-operator.md) | WeChat private traffic empire. |
+| 📱 [WeChat Official Account Manager](marketing/china-market/wechat-official-account.md) | Loyal WeChat subscriber communities. |
+| 🔥 [Weibo Strategist](marketing/china-market/weibo-strategist.md) | Makes your brand trend on Weibo. |
+| 🌸 [Xiaohongshu Specialist](marketing/china-market/xiaohongshu-specialist.md) | Lifestyle and aesthetic storytelling on 小红书. |
+| 🧠 [Zhihu Strategist](marketing/china-market/zhihu-strategist.md) | Brand authority through expert Q&A. |
+
+#### [`marketing/content`](marketing/content/) — Content, books, growth
+
+| Agent | Vibe |
+|-------|------|
+| 📘 [Book Co-Author](marketing/content/book-co-author.md) | Expertise → book people quote and buy into. |
+| ✍️ [Content Creator](marketing/content/content-creator.md) | Stories across every platform your audience uses. |
+| 🚀 [Growth Hacker](marketing/content/growth-hacker.md) | Unexploited channel — find and scale. |
+
+#### [`marketing/ecommerce`](marketing/ecommerce/) — Cross-border & livestream commerce
+
+| Agent | Vibe |
+|-------|------|
+| 🌏 [Cross-Border E-Commerce Specialist](marketing/ecommerce/cross-border-ecommerce.md) | Chinese factory → global bestseller. |
+| 🎙️ [Livestream Commerce Coach](marketing/ecommerce/livestream-commerce-coach.md) | Hosts: awkward beginner → million-yuan seller. |
+
+#### [`marketing/seo`](marketing/seo/) — Search, AI citations, app store
+
+| Agent | Vibe |
+|-------|------|
+| 🤖 [Agentic Search Optimizer](marketing/seo/agentic-search-optimizer.md) | Makes sure AI can actually do the thing on your site. |
+| 🔮 [AI Citation Strategist](marketing/seo/ai-citation-strategist.md) | Rewires signals so AI recommends you, not the competitor. |
+| 📱 [App Store Optimizer](marketing/seo/app-store-optimizer.md) | Found, downloaded, loved in the store. |
+| 🇨🇳 [Baidu SEO Specialist](marketing/seo/baidu-seo-specialist.md) | Ranks in China's search ecosystem. |
+| 🔍 [SEO Specialist](marketing/seo/seo-specialist.md) | Technical SEO + content for sustainable organic. |
+| 🎬 [Video Optimization Specialist](marketing/seo/video-optimization-specialist.md) | Data-driven retention obsession. |
+
+#### [`marketing/social-media`](marketing/social-media/) — TikTok, Instagram, LinkedIn, Reddit, X
+
+| Agent | Vibe |
+|-------|------|
+| 🎠 [Carousel Growth Engine](marketing/social-media/carousel-growth-engine.md) | Viral carousels from any URL, auto-published. |
+| 📸 [Instagram Curator](marketing/social-media/instagram-curator.md) | Grid aesthetic → engaged community. |
+| 💼 [LinkedIn Content Creator](marketing/social-media/linkedin-content-creator.md) | Scroll-stopping professional content. |
+| 🎧 [Podcast Strategist](marketing/social-media/podcast-strategist.md) | Podcast from concept to loyal audience. |
+| 💬 [Reddit Community Builder](marketing/social-media/reddit-community-builder.md) | Fluent Reddit. Trust the authentic way. |
+| 🎬 [Short-Video Editing Coach](marketing/social-media/short-video-editing-coach.md) | Raw footage → scroll-stopping short videos. |
+| 📣 [Social Media Strategist](marketing/social-media/social-media-strategist.md) | Cross-platform campaigns that build community. |
+| 🎵 [TikTok Strategist](marketing/social-media/tiktok-strategist.md) | Rides the algorithm, builds the culture. |
+| 🐦 [Twitter Engager](marketing/social-media/twitter-engager.md) | Thought leadership 280 chars at a time. |
+
+### Paid Media
+
+#### [`paid-media/analytics`](paid-media/analytics/) — Audits, attribution, search-term mining
+
+| Agent | Vibe |
+|-------|------|
+| 📋 [Paid Media Auditor](paid-media/analytics/auditor.md) | Finds the waste before your CFO does. |
+| 🔍 [Search Query Analyst](paid-media/analytics/search-query-analyst.md) | Mines queries for gold competitors miss. |
+| 📡 [Tracking & Measurement Specialist](paid-media/analytics/tracking-specialist.md) | If it's not tracked correctly, it didn't happen. |
+
+#### [`paid-media/channels`](paid-media/channels/) — Creative, PPC, paid social, programmatic
+
+| Agent | Vibe |
+|-------|------|
+| ✍️ [Ad Creative Strategist](paid-media/channels/creative-strategist.md) | Creative as a repeatable science. |
+| 📱 [Paid Social Strategist](paid-media/channels/paid-social-strategist.md) | Meta, LinkedIn, TikTok — every dollar working harder. |
+| 💰 [PPC Campaign Strategist](paid-media/channels/ppc-strategist.md) | $10K to $10M+/mo PPC architectures. |
+| 📺 [Programmatic & Display Buyer](paid-media/channels/programmatic-buyer.md) | Display and video at scale. |
+
+### Product
+
+#### [`product/delivery`](product/delivery/) — PM, prioritization, behavioral
+
+| Agent | Vibe |
+|-------|------|
+| 🧠 [Behavioral Nudge Engine](product/delivery/behavioral-nudge-engine.md) | Behavioral psych for user motivation. |
+| 🧭 [Product Manager](product/delivery/manager.md) | Ships the right thing, not the next thing. |
+| 🎯 [Sprint Prioritizer](product/delivery/sprint-prioritizer.md) | Ruthless focus on sprint value. |
+
+#### [`product/discovery`](product/discovery/) — Research & feedback
+
+| Agent | Vibe |
+|-------|------|
+| 🔍 [Feedback Synthesizer](product/discovery/feedback-synthesizer.md) | 1000 user voices → 5 things to build. |
+| 🔭 [Trend Researcher](product/discovery/trend-researcher.md) | Trends before mainstream. |
+
+### Project Management
+
+#### [`project-management/delivery`](project-management/delivery/) — Experiments & Jira workflow
+
+| Agent | Vibe |
+|-------|------|
+| 🧪 [Experiment Tracker](project-management/delivery/experiment-tracker.md) | Designs experiments, tracks results, lets data decide. |
+| 📋 [Jira Workflow Steward](project-management/delivery/jira-workflow-steward.md) | Traceable commits, structured PRs, release-safe branches. |
+
+#### [`project-management/planning`](project-management/planning/) — PM, scoping, shepherding
+
+| Agent | Vibe |
+|-------|------|
+| 🐑 [Project Shepherd](project-management/planning/project-shepherd.md) | Cross-functional chaos → on-time delivery. |
+| 📝 [Senior Project Manager](project-management/planning/project-manager-senior.md) | Realistic scope — no gold-plating. |
+
+#### [`project-management/studio`](project-management/studio/) — Studio production & ops
+
+| Agent | Vibe |
+|-------|------|
+| 🏭 [Studio Operations](project-management/studio/studio-operations.md) | Processes, tools, people in sync. |
+| 🎬 [Studio Producer](project-management/studio/studio-producer.md) | Creative vision aligned to business. |
+
+### Sales
+
+#### [`sales/coaching`](sales/coaching/) — Discovery and rep coaching
+
+| Agent | Vibe |
+|-------|------|
+| 🔍 [Discovery Coach](sales/coaching/discovery-coach.md) | The one extra question that closes the deal. |
+| 🏋️ [Sales Coach](sales/coaching/coach.md) | Makes the rep rethink the deal. |
+
+#### [`sales/ops`](sales/ops/) — Pipeline analytics & sales engineering
 
-Content files are discovered recursively. Each level can have its own `bank-template.json` with a tailored mission and knowledge pages.
+| Agent | Vibe |
+|-------|------|
+| 📊 [Pipeline Analyst](sales/ops/pipeline-analyst.md) | Tells you your forecast is wrong before you do. |
+| 🛠️ [Sales Engineer](sales/ops/engineer.md) | Wins the technical decision pre-procurement. |
 
-## What `install` does
+#### [`sales/strategy`](sales/strategy/) — Account, deal, outbound, proposal
 
-The CLI is a single setup command — no manual steps required. Here's what happens:
+| Agent | Vibe |
+|-------|------|
+| 🗺️ [Account Strategist](sales/strategy/account-strategist.md) | Maps the org, finds the whitespace, expands. |
+| ♟️ [Deal Strategist](sales/strategy/deal-strategist.md) | Qualifies like a surgeon, kills happy ears. |
+| 🎯 [Outbound Strategist](sales/strategy/outbound-strategist.md) | Buying signals into booked meetings. |
+| 🏹 [Proposal Strategist](sales/strategy/proposal-strategist.md) | RFP responses buyers can't put down. |
 
-1. **Fetches the agent** — downloads the directory from GitHub (or uses a local path)
-2. **Configures connection** — prompts for Hindsight Cloud or self-hosted API URL and token
-3. **Connects to Hindsight** and imports the `bank-template.json` (memory bank config, knowledge pages, directives)
-4. **Ingests all content files** (`.md`, `.txt`, etc.) found recursively as seed knowledge
-5. **Generates the agent** — creates a skill (Claude) or workspace (OpenClaw/NemoClaw) with the knowledge tools baked in
+### Spatial Computing
 
-### Claude harness
+#### [`spatial-computing/platform`](spatial-computing/platform/) — visionOS, Metal, terminals
 
-For Claude Chat and Cowork, the CLI generates a self-contained skill zip:
+| Agent | Vibe |
+|-------|------|
+| 🍎 [macOS Spatial/Metal Engineer](spatial-computing/platform/macos-spatial-metal-engineer.md) | Metal pushed to its limits on macOS and Vision Pro. |
+| 🖥️ [Terminal Integration Specialist](spatial-computing/platform/terminal-integration-specialist.md) | Terminal emulation and text rendering in Swift. |
+| 🥽 [visionOS Spatial Engineer](spatial-computing/platform/visionos-spatial-engineer.md) | Native volumetric interfaces, Liquid Glass. |
 
-- The agent's Hindsight API URL, bank ID, and token are baked directly into the skill
-- No external dependencies — uses `curl` to call the Hindsight REST API
-- The skill instructs Claude to load knowledge pages at startup, retain user feedback during conversations, and manage pages
-- Upload the zip via **Customize → Skills → Upload** in Claude Chat or Cowork
-- After upload, allowlist the API host in **Settings → Capabilities**
-- Type `/<agent-name>` in any conversation to activate
+#### [`spatial-computing/xr`](spatial-computing/xr/) — XR/WebXR/cockpit
 
-### OpenClaw / NemoClaw harness
+| Agent | Vibe |
+|-------|------|
+| 🕹️ [XR Cockpit Interaction Specialist](spatial-computing/xr/xr-cockpit-interaction-specialist.md) | Immersive cockpit control systems. |
+| 🌐 [XR Immersive Developer](spatial-computing/xr/xr-immersive-developer.md) | Browser-based AR/VR/XR pushing WebXR. |
+| 🫧 [XR Interface Architect](spatial-computing/xr/xr-interface-architect.md) | Spatial interaction that feels like instinct. |
+
+### Specialized
+
+#### [`specialized/automation`](specialized/automation/) — Data, docs, distribution
+
+| Agent | Vibe |
+|-------|------|
+| 🗄️ [Data Consolidation Agent](specialized/automation/data-consolidation-agent.md) | Scattered sales data → live dashboards. |
+| 📄 [Document Generator](specialized/automation/document-generator.md) | PDFs, slides, spreadsheets from code. |
+| 📤 [Report Distribution Agent](specialized/automation/report-distribution-agent.md) | Reports delivered to the right reps. |
+
+#### [`specialized/compliance-trust`](specialized/compliance-trust/) — Identity, audits, ZK
+
+| Agent | Vibe |
+|-------|------|
+| 🔐 [Agentic Identity & Trust Architect](specialized/compliance-trust/agentic-identity-trust.md) | Every agent proves who, what, when. |
+| ⚙️ [Automation Governance Architect](specialized/compliance-trust/automation-governance-architect.md) | Reliable systems > automation hype. |
+| 🛡️ [Blockchain Security Auditor](specialized/compliance-trust/blockchain-security-auditor.md) | Finds the exploit before the attacker. |
+| 📋 [Compliance Auditor](specialized/compliance-trust/compliance-auditor.md) | Readiness → SOC 2 certification. |
+| 🕸️ [Identity Graph Operator](specialized/compliance-trust/identity-graph-operator.md) | Same canonical answer across all agents. |
+| 🗃️ [ZK Steward](specialized/compliance-trust/zk-steward.md) | Zettelkasten for validated knowledge bases. |
+
+#### [`specialized/customer-service`](specialized/customer-service/) — Hospitality, retail, general
+
+| Agent | Vibe |
+|-------|------|
+| 🎧 [Customer Service](specialized/customer-service/customer-service.md) | Problem → loyalty, with speed and care. |
+| 🏨 [Hospitality Guest Services](specialized/customer-service/hospitality-guest-services.md) | Hospitality is a feeling. |
+| 🛒 [Retail Customer Returns](specialized/customer-service/retail-customer-returns.md) | A return is an opportunity. |
+
+#### [`specialized/domain-experts`](specialized/domain-experts/) — Civil, devrel, gov, supply chain
+
+| Agent | Vibe |
+|-------|------|
+| 🏗️ [Civil Engineer](specialized/domain-experts/civil-engineer.md) | Structures from seismic Tokyo to wind-swept Dubai. |
+| 🗣️ [Developer Advocate](specialized/domain-experts/developer-advocate.md) | Product team ↔ developer community. |
+| 🏛️ [Government Digital Presales Consultant](specialized/domain-experts/government-digital-presales-consultant.md) | China gov IT procurement, navigated. |
+| 🔗 [Supply Chain Strategist](specialized/domain-experts/supply-chain-strategist.md) | Procurement and resilience across China manufacturing. |
+
+#### [`specialized/engineering-meta`](specialized/engineering-meta/) — LSP, MCP, model QA, Salesforce, workflow
+
+| Agent | Vibe |
+|-------|------|
+| 🔎 [LSP/Index Engineer](specialized/engineering-meta/lsp-index-engineer.md) | LSP orchestration + semantic indexing. |
+| 🔌 [MCP Builder](specialized/engineering-meta/mcp-builder.md) | Tools that make AI agents useful. |
+| 🔬 [Model QA Specialist](specialized/engineering-meta/model-qa.md) | ML audits end-to-end. |
+| ☁️ [Salesforce Architect](specialized/engineering-meta/salesforce-architect.md) | Tangled org → architecture that scales. |
+| 🗺️ [Workflow Architect](specialized/engineering-meta/workflow-architect.md) | Every path mapped before code is written. |
+
+#### [`specialized/exec-ops`](specialized/exec-ops/) — Orchestration & chief of staff
+
+| Agent | Vibe |
+|-------|------|
+| 🎛️ [Agents Orchestrator](specialized/exec-ops/agents-orchestrator.md) | Conducts the entire dev pipeline. |
+| 🧭 [Chief of Staff](specialized/exec-ops/chief-of-staff.md) | I own the space between functions. |
+
+#### [`specialized/finance-ops`](specialized/finance-ops/) — AP, loans, real estate
+
+| Agent | Vibe |
+|-------|------|
+| 💸 [Accounts Payable Agent](specialized/finance-ops/accounts-payable-agent.md) | Crypto, fiat, stablecoins — any rail. |
+| 🏦 [Loan Officer Assistant](specialized/finance-ops/loan-officer-assistant.md) | Pipeline with precision and compliance. |
+| 🏠 [Real Estate Buyer & Seller](specialized/finance-ops/real-estate-buyer-seller.md) | Biggest financial decision, handled. |
+
+#### [`specialized/healthcare`](specialized/healthcare/) — Patient service & compliance
+
+| Agent | Vibe |
+|-------|------|
+| 🏥 [Healthcare Customer Service](specialized/healthcare/healthcare-customer-service.md) | Patients heard, respected, supported. |
+| ⚕️ [Healthcare Marketing Compliance](specialized/healthcare/healthcare-marketing-compliance.md) | Legal in China's healthcare landscape. |
+
+#### [`specialized/hr`](specialized/hr/) — Training, onboarding, recruiting
+
+| Agent | Vibe |
+|-------|------|
+| 📚 [Corporate Training Designer](specialized/hr/corporate-training-designer.md) | Training that drives behavior change. |
+| 🤝 [HR Onboarding](specialized/hr/hr-onboarding.md) | First 90 days decide long-term outcome. |
+| 🎯 [Recruitment Specialist](specialized/hr/recruitment-specialist.md) | Full-cycle recruiting across hiring platforms. |
+
+#### [`specialized/legal`](specialized/legal/) — Intake, review, billing
+
+| Agent | Vibe |
+|-------|------|
+| ⏱️ [Legal Billing & Time Tracking](specialized/legal/legal-billing-time-tracking.md) | Six minutes unbilled is money on the table. |
+| 📋 [Legal Client Intake](specialized/legal/legal-client-intake.md) | First conversation sets the tone. |
+| ⚖️ [Legal Document Review](specialized/legal/legal-document-review.md) | Every word matters; every missed clause is liability. |
+
+#### [`specialized/regional`](specialized/regional/) — Cross-cultural, French, Korean, study abroad, translation
+
+| Agent | Vibe |
+|-------|------|
+| 🌍 [Cultural Intelligence Strategist](specialized/regional/cultural-intelligence-strategist.md) | Detects invisible exclusion across cultures. |
+| 🇫🇷 [French Consulting Market Navigator](specialized/regional/french-consulting-market.md) | Decodes the French consulting food chain. |
+| 🇰🇷 [Korean Business Navigator](specialized/regional/korean-business-navigator.md) | Western directness ↔ Korean relationship dynamics. |
+| 🌐 [Language Translator](specialized/regional/language-translator.md) | Precision with cultural respect. |
+| 🎓 [Study Abroad Advisor](specialized/regional/study-abroad-advisor.md) | School, essays, visas — end to end. |
+
+#### [`specialized/sales-ops`](specialized/sales-ops/) — Extraction & outreach
+
+| Agent | Vibe |
+|-------|------|
+| 📊 [Sales Data Extraction Agent](specialized/sales-ops/sales-data-extraction-agent.md) | Watches Excel, extracts the metrics. |
+| 🎯 [Sales Outreach](specialized/sales-ops/sales-outreach.md) | Help people buy, don't sell. |
+
+### Support
+
+#### [`support/analytics`](support/analytics/) — Reporting, finance, executive summaries
+
+| Agent | Vibe |
+|-------|------|
+| 📊 [Analytics Reporter](support/analytics/analytics-reporter.md) | Raw data → insights that drive the next decision. |
+| 📝 [Executive Summary Generator](support/analytics/executive-summary-generator.md) | Thinks McKinsey, writes for the C-suite. |
+| 💰 [Finance Tracker](support/analytics/finance-tracker.md) | Clean books, honest forecasts. |
+
+#### [`support/ops-response`](support/ops-response/) — Infra, compliance, support
+
+| Agent | Vibe |
+|-------|------|
+| 🏢 [Infrastructure Maintainer](support/ops-response/infrastructure-maintainer.md) | Lights on, servers humming, alerts quiet. |
+| ⚖️ [Legal Compliance Checker](support/ops-response/legal-compliance-checker.md) | Compliance across every jurisdiction. |
+| 💬 [Support Responder](support/ops-response/support-responder.md) | Frustrated users → loyal advocates. |
+
+### Testing
+
+#### [`testing/automation`](testing/automation/) — APIs, evidence, results
+
+| Agent | Vibe |
+|-------|------|
+| 🔌 [API Tester](testing/automation/api-tester.md) | Breaks your API before users do. |
+| 📸 [Evidence Collector](testing/automation/evidence-collector.md) | Won't approve without visual proof. |
+| 📋 [Test Results Analyzer](testing/automation/test-results-analyzer.md) | Reads results like a detective reads evidence. |
+
+#### [`testing/meta`](testing/meta/) — Tool evaluation & workflow tuning
+
+| Agent | Vibe |
+|-------|------|
+| 🔧 [Tool Evaluator](testing/meta/tool-evaluator.md) | The right tools, recommended. |
+| ⚡ [Workflow Optimizer](testing/meta/workflow-optimizer.md) | Bottleneck → fixed → automated. |
+
+#### [`testing/quality`](testing/quality/) — Accessibility, performance, reality
+
+| Agent | Vibe |
+|-------|------|
+| ♿ [Accessibility Auditor](testing/quality/accessibility-auditor.md) | If it's not screen-reader tested, it's not accessible. |
+| ⏱️ [Performance Benchmarker](testing/quality/performance-benchmarker.md) | Measures, optimizes, proves it. |
+| 🧐 [Reality Checker](testing/quality/reality-checker.md) | Defaults to "NEEDS WORK" — overwhelming proof required. |
+
+---
+
+## Build your own
+
+An agent is a directory:
+
+```
+my-agent/
+  bank-template.json     # memory bank config (optional)
+  *.md                   # any markdown becomes seed knowledge
+```
 
-For OpenClaw and NemoClaw, the CLI installs the Hindsight plugin, creates a workspace, and registers the agent with the harness. Restart the gateway and start chatting.
+Nest for multi-level teams. Each level can have its own `bank-template.json`. Install with any of the local/repo/blank patterns shown at the top.
 
-## Supported harnesses
+## Harnesses
 
-| Harness | Flag | Status |
-|---------|------|--------|
-| [Claude Chat](https://claude.ai) | `--harness claude` | Supported |
-| [Claude Cowork](https://claude.ai) | `--harness claude` | Supported |
-| [OpenClaw](https://openclaw.dev) | `--harness openclaw` | Supported |
-| [NemoClaw](https://github.com/NVIDIA/NeMo-Agent) | `--harness nemoclaw` | Supported |
+| Harness | Flag |
+|---------|------|
+| [Claude Chat](https://claude.ai) / Cowork | `--harness claude` |
+| [Claude Code](https://claude.com/claude-code) | `--harness claude-code` |
+| [OpenClaw](https://openclaw.dev) | `--harness openclaw` |
+| [NemoClaw](https://github.com/NVIDIA/NeMo-Agent) | `--harness nemoclaw` |
+| Hermes | `--harness hermes` |
 
-## Requirements
+## License
 
-- A supported harness
-- A Hindsight account ([Cloud](https://hindsight.vectorize.io) or self-hosted)
-- Everything else is handled by the CLI
+MIT.

package/dist/cli.js

@@ -660,7 +660,9 @@ curl -s -X POST ${apiUrl}/v1/default/banks/${bankId}/memories/retain \\
     execSync(`cd "${outDir}" && zip -j "${zipPath}" SKILL.md`, { stdio: "pipe" });
     return zipPath;
 }
-async function promptClaudeConfig(agentId, opts = { askBankId: true }) {
+async function promptClaudeConfig(agentId, opts = {}) {
+    const askBankId = opts.askBankId ?? true;
+    const allowLocalhost = opts.allowLocalhost ?? false;
     const deploymentType = await p.select({
         message: "Hindsight deployment:",
         options: [
@@ -679,11 +681,14 @@ async function promptClaudeConfig(agentId, opts = { askBankId: true }) {
     else {
         const urlInput = await p.text({
             message: "Hindsight API URL:",
-            placeholder: "https://your-hindsight.example.com",
+            placeholder: allowLocalhost
+                ? "http://localhost:9077 or https://your-hindsight.example.com"
+                : "https://your-hindsight.example.com",
             validate: (val) => {
                 if (!val)
                     return "URL is required";
-                if (val.startsWith("http://localhost") || val.startsWith("http://127.0.0.1")) {
+                if (!allowLocalhost &&
+                    (val.startsWith("http://localhost") || val.startsWith("http://127.0.0.1"))) {
                     return "Claude cannot reach localhost. Use a publicly accessible URL.";
                 }
                 return undefined;
@@ -694,7 +699,9 @@ async function promptClaudeConfig(agentId, opts = { askBankId: true }) {
             process.exit(0);
         }
         apiUrl = urlInput;
-        p.log.warn("Make sure your Hindsight instance is publicly accessible from Claude's servers.");
+        if (!allowLocalhost) {
+            p.log.warn("Make sure your Hindsight instance is publicly accessible from Claude's servers.");
+        }
     }
     const tokenInput = await p.password({ message: "Hindsight API token:" });
     if (p.isCancel(tokenInput)) {
@@ -703,7 +710,7 @@ async function promptClaudeConfig(agentId, opts = { askBankId: true }) {
     }
     const apiToken = tokenInput || undefined;
     let bankId = agentId;
-    if (opts.askBankId) {
+    if (askBankId) {
         const bankInput = await p.text({
             message: "Bank ID:",
             initialValue: agentId,
@@ -992,9 +999,29 @@ async function main() {
                 catch { /* ignore */ }
             }
             const hasConnection = ccConfig.hindsightApiUrl || ccConfig.llmProvider;
-            if (!hasConnection && process.stdin.isTTY) {
+            let reconfigure = !hasConnection;
+            if (hasConnection && process.stdin.isTTY) {
+                const summary = ccConfig.hindsightApiUrl === HINDSIGHT_CLOUD_API_URL
+                    ? "Cloud (api.hindsight.vectorize.io)"
+                    : ccConfig.hindsightApiUrl
+                        ? `External: ${ccConfig.hindsightApiUrl}`
+                        : `LLM provider: ${ccConfig.llmProvider}`;
+                const ok = await p.confirm({
+                    message: `Hindsight: ${color.cyan(summary)}. Use this?\n${color.dim("  Changing this will affect all existing Claude Code agents — one Claude Code install shares a single Hindsight connection.")}`,
+                });
+                if (p.isCancel(ok)) {
+                    p.cancel("Cancelled.");
+                    process.exit(0);
+                }
+                if (!ok)
+                    reconfigure = true;
+            }
+            if (reconfigure && process.stdin.isTTY) {
                 // Bank ID is derived from agent + cwd at runtime — don't ask for it
-                const claudeConfig = await promptClaudeConfig(agentId, { askBankId: false });
+                const claudeConfig = await promptClaudeConfig(agentId, {
+                    askBankId: false,
+                    allowLocalhost: true,
+                });
                 ccConfig.hindsightApiUrl = claudeConfig.apiUrl;
                 ccConfig.hindsightApiToken = claudeConfig.apiToken;
             }

package/dist/skill/SKILL.md

@@ -41,7 +41,9 @@ Use when pages don't cover what you need.
 
 ## Ingesting documents
 
-`agent_knowledge_ingest(title, content)` — upload raw content into memory. Never summarize before ingesting. Save large content to a file first, read it, then pass the full text.
+`agent_knowledge_ingest(title, content)` — upload raw content into memory. Never summarize before ingesting. Pass the full text inline.
+
+`agent_knowledge_ingest_files(paths)` — ingest one or more files straight from disk. `paths` is a list of file paths or glob patterns (e.g. `["docs/**/*.md", "/abs/path/notes.txt"]`). Each file's content is read and stored under a document ID derived from its path. Prefer this over `agent_knowledge_ingest` when the content already lives in files — no need to read them first. Use absolute paths when in doubt; relative paths resolve against the working directory.
 
 ## Updating and deleting
 

package/dist/tests/lint.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/lint.test.js

@@ -0,0 +1,85 @@
+import { describe, it, expect, beforeEach, afterEach } from "vitest";
+import { mkdtempSync, rmSync, writeFileSync, mkdirSync } from "fs";
+import { tmpdir } from "os";
+import { join, resolve } from "path";
+// @ts-expect-error - .mjs has no type declarations; runtime import is fine
+import { findBankTemplates, lintBankTemplate, lintAll } from "../scripts/lint-bank-templates.mjs";
+const REPO_ROOT = resolve(__dirname, "..", "..");
+describe("lint-bank-templates", () => {
+    let dir;
+    beforeEach(() => {
+        dir = mkdtempSync(join(tmpdir(), "sda-lint-"));
+    });
+    afterEach(() => {
+        rmSync(dir, { recursive: true, force: true });
+    });
+    it("accepts a valid template (observations_mission + >=1 mental_models, no reflect_mission)", () => {
+        const file = join(dir, "bank-template.json");
+        writeFileSync(file, JSON.stringify({
+            version: "1",
+            bank: { observations_mission: "watch X" },
+            mental_models: [{ id: "a", name: "A" }],
+        }));
+        expect(lintBankTemplate(file)).toEqual([]);
+    });
+    it("rejects missing observations_mission", () => {
+        const file = join(dir, "bank-template.json");
+        writeFileSync(file, JSON.stringify({ bank: {}, mental_models: [{ id: "a", name: "A" }] }));
+        const errs = lintBankTemplate(file);
+        expect(errs.some((e) => e.includes("observations_mission"))).toBe(true);
+    });
+    it("rejects empty observations_mission", () => {
+        const file = join(dir, "bank-template.json");
+        writeFileSync(file, JSON.stringify({
+            bank: { observations_mission: "   " },
+            mental_models: [{ id: "a", name: "A" }],
+        }));
+        const errs = lintBankTemplate(file);
+        expect(errs.some((e) => e.includes("observations_mission"))).toBe(true);
+    });
+    it("rejects empty mental_models", () => {
+        const file = join(dir, "bank-template.json");
+        writeFileSync(file, JSON.stringify({
+            bank: { observations_mission: "watch X" },
+            mental_models: [],
+        }));
+        const errs = lintBankTemplate(file);
+        expect(errs.some((e) => e.includes("mental_models"))).toBe(true);
+    });
+    it("rejects missing mental_models entirely", () => {
+        const file = join(dir, "bank-template.json");
+        writeFileSync(file, JSON.stringify({ bank: { observations_mission: "watch X" } }));
+        const errs = lintBankTemplate(file);
+        expect(errs.some((e) => e.includes("mental_models"))).toBe(true);
+    });
+    it("rejects deprecated reflect_mission", () => {
+        const file = join(dir, "bank-template.json");
+        writeFileSync(file, JSON.stringify({
+            bank: { observations_mission: "watch X", reflect_mission: "old" },
+            mental_models: [{ id: "a", name: "A" }],
+        }));
+        const errs = lintBankTemplate(file);
+        expect(errs.some((e) => e.includes("reflect_mission"))).toBe(true);
+    });
+    it("rejects invalid JSON", () => {
+        const file = join(dir, "bank-template.json");
+        writeFileSync(file, "{ not json");
+        const errs = lintBankTemplate(file);
+        expect(errs.some((e) => e.toLowerCase().includes("json"))).toBe(true);
+    });
+    it("findBankTemplates walks recursively and skips node_modules", () => {
+        mkdirSync(join(dir, "a", "b"), { recursive: true });
+        mkdirSync(join(dir, "node_modules", "x"), { recursive: true });
+        writeFileSync(join(dir, "bank-template.json"), "{}");
+        writeFileSync(join(dir, "a", "bank-template.json"), "{}");
+        writeFileSync(join(dir, "a", "b", "bank-template.json"), "{}");
+        writeFileSync(join(dir, "node_modules", "x", "bank-template.json"), "{}");
+        const found = findBankTemplates(dir);
+        expect(found).toHaveLength(3);
+        expect(found.every((f) => !f.includes("node_modules"))).toBe(true);
+    });
+    it("the actual repo passes lint", () => {
+        const issues = lintAll(REPO_ROOT);
+        expect(issues).toEqual([]);
+    });
+});

package/hermes-plugin/__init__.py

@@ -0,0 +1,470 @@
+"""Self-driving agents Hindsight plugin for Hermes.
+
+Registers agent_knowledge_* tools as a regular plugin (not a memory provider),
+so it coexists with the bundled hindsight memory provider or any other provider.
+
+Config read from the active Hermes profile's hindsight/config.json (the same file
+the bundled hindsight provider uses), so both share one bank, API URL, and key:
+  { "api_url": "...", "api_key": "...", "bank_id": "..." }
+"""
+
+from __future__ import annotations
+
+import glob
+import json
+import logging
+import os
+from pathlib import Path
+from typing import Any
+
+import httpx
+
+logger = logging.getLogger(__name__)
+
+PAGE_DEFAULTS = {
+    "mode": "delta",
+    "refresh_after_consolidation": True,
+    "exclude_mental_models": True,
+    "fact_types": ["observation"],
+}
+
+
+def _get_hermes_home() -> Path:
+    """Resolve the active Hermes home, profile-isolation aware.
+
+    Prefer Hermes's own resolver: it honors the in-process profile override and
+    the active-profile marker, not just the ``HERMES_HOME`` env var. This is what
+    makes us read the *same* config.json the bundled hindsight provider reads —
+    if we only checked the env var (which is often unset in the agent worker) we
+    could silently fall back to the global ~/.hermes config and write to the
+    wrong bank. Fall back to the env var, then the default, if Hermes isn't importable.
+    """
+    try:
+        from hermes_constants import get_hermes_home  # type: ignore
+        return Path(get_hermes_home())
+    except Exception:
+        env = os.environ.get("HERMES_HOME")
+        if env:
+            return Path(env)
+        return Path.home() / ".hermes"
+
+
+def _load_config() -> dict | None:
+    """Load Hindsight config from the active Hermes profile.
+
+    Reads the same config.json the bundled hindsight provider uses,
+    so both share the same bank, API URL, and credentials.
+    """
+    cfg_path = _get_hermes_home() / "hindsight" / "config.json"
+    if not cfg_path.exists():
+        return None
+    try:
+        cfg = json.loads(cfg_path.read_text())
+        # Normalize field names (bundled provider uses api_url/api_key,
+        # we expose as api_url/api_token for consistency with other harnesses)
+        return {
+            "api_url": cfg.get("api_url", ""),
+            "api_token": cfg.get("api_key", ""),
+            "bank_id": cfg.get("bank_id", "hermes"),
+        }
+    except Exception:
+        return None
+
+
+def _api(
+    api_url: str,
+    path: str,
+    method: str = "GET",
+    body: dict | None = None,
+    token: str | None = None,
+    timeout: float = 30.0,
+) -> Any:
+    headers: dict[str, str] = {"Content-Type": "application/json"}
+    if token:
+        headers["Authorization"] = f"Bearer {token}"
+    resp = httpx.request(
+        method,
+        f"{api_url}{path}",
+        json=body,
+        headers=headers,
+        timeout=timeout,
+    )
+    resp.raise_for_status()
+    return resp.json() if resp.content else {}
+
+
+def _is_available() -> bool:
+    return _load_config() is not None
+
+
+# ── Tool handlers ───────────────────────────────────────
+
+
+def _handle_list_pages(args: dict, **kwargs: Any) -> str:
+    config = _load_config()
+    if not config:
+        return json.dumps({"error": "Plugin not configured"})
+    api_url = config["api_url"].rstrip("/")
+    token = config.get("api_token")
+    bank_id = config["bank_id"]
+    try:
+        result = _api(api_url, f"/v1/default/banks/{bank_id}/mental-models?detail=metadata", "GET", token=token)
+        return json.dumps(result, indent=2)
+    except Exception as e:
+        return json.dumps({"error": str(e)})
+
+
+def _handle_get_page(args: dict, **kwargs: Any) -> str:
+    config = _load_config()
+    if not config:
+        return json.dumps({"error": "Plugin not configured"})
+    api_url = config["api_url"].rstrip("/")
+    token = config.get("api_token")
+    bank_id = config["bank_id"]
+    try:
+        result = _api(api_url, f"/v1/default/banks/{bank_id}/mental-models/{args['page_id']}", "GET", token=token)
+        return json.dumps(result, indent=2)
+    except Exception as e:
+        return json.dumps({"error": str(e)})
+
+
+def _handle_create_page(args: dict, **kwargs: Any) -> str:
+    config = _load_config()
+    if not config:
+        return json.dumps({"error": "Plugin not configured"})
+    api_url = config["api_url"].rstrip("/")
+    token = config.get("api_token")
+    bank_id = config["bank_id"]
+    try:
+        result = _api(
+            api_url,
+            f"/v1/default/banks/{bank_id}/mental-models",
+            "POST",
+            body={
+                "id": args["page_id"],
+                "name": args["name"],
+                "source_query": args["source_query"],
+                "max_tokens": 4096,
+                "trigger": PAGE_DEFAULTS,
+            },
+            token=token,
+        )
+        return json.dumps(result, indent=2)
+    except Exception as e:
+        return json.dumps({"error": str(e)})
+
+
+def _handle_update_page(args: dict, **kwargs: Any) -> str:
+    config = _load_config()
+    if not config:
+        return json.dumps({"error": "Plugin not configured"})
+    api_url = config["api_url"].rstrip("/")
+    token = config.get("api_token")
+    bank_id = config["bank_id"]
+    try:
+        body: dict[str, str] = {}
+        if args.get("name"):
+            body["name"] = args["name"]
+        if args.get("source_query"):
+            body["source_query"] = args["source_query"]
+        result = _api(api_url, f"/v1/default/banks/{bank_id}/mental-models/{args['page_id']}", "PATCH", body=body, token=token)
+        return json.dumps(result, indent=2)
+    except Exception as e:
+        return json.dumps({"error": str(e)})
+
+
+def _handle_delete_page(args: dict, **kwargs: Any) -> str:
+    config = _load_config()
+    if not config:
+        return json.dumps({"error": "Plugin not configured"})
+    api_url = config["api_url"].rstrip("/")
+    token = config.get("api_token")
+    bank_id = config["bank_id"]
+    try:
+        _api(api_url, f"/v1/default/banks/{bank_id}/mental-models/{args['page_id']}", "DELETE", token=token)
+        return json.dumps({"success": True})
+    except Exception as e:
+        return json.dumps({"error": str(e)})
+
+
+def _handle_recall(args: dict, **kwargs: Any) -> str:
+    config = _load_config()
+    if not config:
+        return json.dumps({"error": "Plugin not configured"})
+    api_url = config["api_url"].rstrip("/")
+    token = config.get("api_token")
+    bank_id = config["bank_id"]
+    try:
+        result = _api(
+            api_url,
+            f"/v1/default/banks/{bank_id}/memories/recall",
+            "POST",
+            body={"query": args["query"], "max_results": args.get("max_results", 10)},
+            token=token,
+        )
+        return json.dumps(result, indent=2)
+    except Exception as e:
+        return json.dumps({"error": str(e)})
+
+
+def _handle_ingest(args: dict, **kwargs: Any) -> str:
+    config = _load_config()
+    if not config:
+        return json.dumps({"error": "Plugin not configured"})
+    api_url = config["api_url"].rstrip("/")
+    token = config.get("api_token")
+    bank_id = config["bank_id"]
+    try:
+        doc_id = args["title"].lower().replace(" ", "-")
+        result = _api(
+            api_url,
+            f"/v1/default/banks/{bank_id}/memories",
+            "POST",
+            body={"items": [{"content": args["content"], "document_id": doc_id}], "async": True},
+            token=token,
+        )
+        return json.dumps(result, indent=2)
+    except Exception as e:
+        return json.dumps({"error": str(e)})
+
+
+def _doc_id_for(path: Path) -> str:
+    """Stable, collision-resistant document id from a file path."""
+    try:
+        rel = path.resolve().relative_to(Path.cwd().resolve())
+    except ValueError:
+        rel = Path(path.name)
+    return str(rel).lower().replace(os.sep, "-").replace(" ", "-").lstrip("-")
+
+
+def _handle_ingest_files(args: dict, **kwargs: Any) -> str:
+    config = _load_config()
+    if not config:
+        return json.dumps({"error": "Plugin not configured"})
+    api_url = config["api_url"].rstrip("/")
+    token = config.get("api_token")
+    bank_id = config["bank_id"]
+
+    patterns = args.get("paths") or []
+    if isinstance(patterns, str):
+        patterns = [patterns]
+    if not patterns:
+        return json.dumps({"error": "No paths provided"})
+
+    # Expand globs (and literal paths), dedupe, keep deterministic order
+    resolved: list[Path] = []
+    seen: set[str] = set()
+    for pat in patterns:
+        expanded = os.path.expanduser(pat)
+        matches = glob.glob(expanded, recursive=True)
+        if not matches and os.path.exists(expanded):
+            matches = [expanded]
+        for m in sorted(matches):
+            p = Path(m)
+            key = str(p.resolve())
+            if p.is_file() and key not in seen:
+                seen.add(key)
+                resolved.append(p)
+
+    if not resolved:
+        return json.dumps({"error": "No files matched", "patterns": patterns})
+
+    items: list[dict] = []
+    errors: list[dict] = []
+    for p in resolved:
+        try:
+            content = p.read_text(encoding="utf-8", errors="replace")
+        except Exception as e:
+            errors.append({"path": str(p), "error": str(e)})
+            continue
+        if not content.strip():
+            errors.append({"path": str(p), "error": "empty file, skipped"})
+            continue
+        items.append({"content": content, "document_id": _doc_id_for(p)})
+
+    if not items:
+        return json.dumps({"error": "No readable, non-empty files", "errors": errors})
+
+    try:
+        result = _api(
+            api_url,
+            f"/v1/default/banks/{bank_id}/memories",
+            "POST",
+            body={"items": items, "async": True},
+            token=token,
+        )
+        return json.dumps(
+            {"ingested": len(items), "document_ids": [it["document_id"] for it in items],
+             "errors": errors, "result": result},
+            indent=2,
+        )
+    except Exception as e:
+        return json.dumps({"error": str(e), "errors": errors})
+
+
+# ── Tool schemas ────────────────────────────────────────
+
+_TOOLS = [
+    (
+        "agent_knowledge_list_pages",
+        {
+            "name": "agent_knowledge_list_pages",
+            "description": "List all your knowledge pages (IDs and names only). Use agent_knowledge_get_page to read the full content of a specific page.",
+            "parameters": {"type": "object", "properties": {}},
+        },
+        _handle_list_pages,
+        "📚",
+    ),
+    (
+        "agent_knowledge_get_page",
+        {
+            "name": "agent_knowledge_get_page",
+            "description": "Read a specific knowledge page by its ID. Returns the full synthesized content.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "page_id": {"type": "string", "description": "Page ID (e.g. 'user-preferences')"},
+                },
+                "required": ["page_id"],
+            },
+        },
+        _handle_get_page,
+        "📖",
+    ),
+    (
+        "agent_knowledge_create_page",
+        {
+            "name": "agent_knowledge_create_page",
+            "description": (
+                "Create a new knowledge page. The source_query is a question the system "
+                "re-asks after each consolidation to rebuild the page from conversation observations."
+            ),
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "page_id": {"type": "string", "description": "Unique page ID, lowercase with hyphens"},
+                    "name": {"type": "string", "description": "Human-readable page name"},
+                    "source_query": {"type": "string", "description": "The question that rebuilds this page"},
+                },
+                "required": ["page_id", "name", "source_query"],
+            },
+        },
+        _handle_create_page,
+        "📝",
+    ),
+    (
+        "agent_knowledge_update_page",
+        {
+            "name": "agent_knowledge_update_page",
+            "description": "Update a page's name or source query. Content re-synthesizes on next consolidation.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "page_id": {"type": "string", "description": "Page ID to update"},
+                    "name": {"type": "string", "description": "New name (optional)"},
+                    "source_query": {"type": "string", "description": "New source query (optional)"},
+                },
+                "required": ["page_id"],
+            },
+        },
+        _handle_update_page,
+        "✏️",
+    ),
+    (
+        "agent_knowledge_delete_page",
+        {
+            "name": "agent_knowledge_delete_page",
+            "description": "Permanently delete a knowledge page.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "page_id": {"type": "string", "description": "Page ID to delete"},
+                },
+                "required": ["page_id"],
+            },
+        },
+        _handle_delete_page,
+        "🗑️",
+    ),
+    (
+        "agent_knowledge_recall",
+        {
+            "name": "agent_knowledge_recall",
+            "description": "Search across all retained conversations and documents for specific facts, numbers, or details not covered by your knowledge pages.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "query": {"type": "string", "description": "What to search for"},
+                    "max_results": {"type": "number", "description": "Max results (default 10)"},
+                },
+                "required": ["query"],
+            },
+        },
+        _handle_recall,
+        "🔍",
+    ),
+    (
+        "agent_knowledge_ingest",
+        {
+            "name": "agent_knowledge_ingest",
+            "description": (
+                "Upload a document into your memory bank. Pass the full raw content — "
+                "never summarize before ingesting. The title becomes the document ID."
+            ),
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "title": {"type": "string", "description": "Document title (becomes the document ID)"},
+                    "content": {"type": "string", "description": "Full raw document content"},
+                },
+                "required": ["title", "content"],
+            },
+        },
+        _handle_ingest,
+        "📥",
+    ),
+    (
+        "agent_knowledge_ingest_files",
+        {
+            "name": "agent_knowledge_ingest_files",
+            "description": (
+                "Ingest one or more files from disk into your memory bank by path. "
+                "Accepts explicit file paths or glob patterns (e.g. 'docs/**/*.md'). "
+                "Reads each file's full raw content — never summarizes. Each file's path "
+                "becomes its document ID. Prefer absolute paths; relative paths resolve "
+                "against the agent's working directory. Use this instead of "
+                "agent_knowledge_ingest when the content already lives in files."
+            ),
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "paths": {
+                        "type": "array",
+                        "items": {"type": "string"},
+                        "description": "File paths or glob patterns to ingest",
+                    },
+                },
+                "required": ["paths"],
+            },
+        },
+        _handle_ingest_files,
+        "📂",
+    ),
+]
+
+
+# ── Registration ────────────────────────────────────────
+
+
+def register(ctx: Any) -> None:
+    """Register agent_knowledge_* tools. Called once by the Hermes plugin loader."""
+    for name, schema, handler, emoji in _TOOLS:
+        ctx.register_tool(
+            name=name,
+            toolset="hindsight-sda",
+            schema=schema,
+            handler=handler,
+            check_fn=_is_available,
+            emoji=emoji,
+        )
+    logger.info("[hindsight-sda] registered %d tools", len(_TOOLS))

package/hermes-plugin/plugin.yaml

@@ -0,0 +1,7 @@
+name: hindsight-sda
+version: 0.2.0
+description: "Self-driving agents knowledge tools powered by Hindsight. Provides agent_knowledge_* tools for managing knowledge pages, recall, and ingestion (raw content or files)."
+kind: standalone
+pip_dependencies:
+  - "httpx>=0.27"
+requires_env: []

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@vectorize-io/self-driving-agents",
-  "version": "0.0.24",
+  "version": "0.0.26",
   "description": "Install self-driving agents with portable memory on any harness",
   "type": "module",
   "main": "dist/cli.js",
@@ -8,11 +8,13 @@
     "self-driving-agents": "dist/cli.js"
   },
   "files": [
-    "dist"
+    "dist",
+    "hermes-plugin"
   ],
   "scripts": {
     "build": "tsc && cp -r src/skill dist/",
     "test": "vitest run src/tests",
+    "lint": "node src/scripts/lint-bank-templates.mjs .",
     "prepublishOnly": "npm run build"
   },
   "keywords": [