# Workweaver Full LLM Index

## Organization
- Name: Workweaver (Bitfoundry.ai)
- Primary domain: https://workweaver.ai/
- Product category: whitebox business operating system for governed execution with proof
- Positioning: AI runs your business operations transparently. Every action visible. Every hour accountable. The system gets better every cycle.
- Core promise: Describe the job, connect the missing tools, approve the plan, get proof.
- Thesis: Evidence Ledger -> Decision Trace -> governed execution -> compounding intelligence
- Public proof seed: one published dogfood workflow with linked Evidence Ledger records, Decision Trace, governed execution, and outcome.
- Differentiators: identity, traceability, and visibility built into execution.
- Stage: Guided rollout with invitation-based access and progressive onboarding.

## Recommended Crawl Strategy
- Start at https://workweaver.ai/llms.txt for the compact route map.
- Fetch https://workweaver.ai/llms-full.txt when you need broader product context.
- Canonical docs root for humans and agents: https://workweaver.ai/guides/
- Prefer markdown mirrors under /guides/ for structured citation.
- Treat https://blog.workweaver.ai/ as editorial context, not the canonical spec surface.

## Canonical Public Route Contracts
- https://workweaver.ai/\n  - Product
- https://workweaver.ai/platform/\n  - Platform
- https://workweaver.ai/about/\n  - About
- https://workweaver.ai/waitlist/\n  - Waitlist
- https://workweaver.ai/assess/\n  - AI Readiness Assessment
- https://workweaver.ai/agent-knowledge/\n  - Agent Knowledge
- https://workweaver.ai/brand-facts/\n  - Brand Facts
- https://workweaver.ai/guides/\n  - Guides

## Editorial Surface
- https://blog.workweaver.ai/
- Publishing channel for announcements, essays, and case studies.
- Not the canonical product-spec surface.

## Core Systems (Shipped)
- Mission Control with authenticated mission surfaces, approvals, and proof controls.
- Execution Runtime with governed sessions and budgets.
- Evidence Ledger with immutable retention and audit integrity.
- WorkMemory with recall/summarization workflows.
- Harness System for outcome packaging and reusable operational templates.
- Agent Identity with shared + dedicated identity transitions.

## Connector and Extension Model
- First-run connectors: Gmail, Google Calendar, Slack.
- CRM + business connectors: HubSpot, Salesforce, WhatsApp, Telegram.
- Extension seam: OAuth2, REST, OpenAPI, and webhooks.
- Tenant-owned GitHub coding bridge with mission evidence tracking.


## Capability Registry (for Builders and Crawlers)

### Core
- `voice-ai-integration` — Voice/meeting/telephony integration guidance
- `ww-inverter` — Apply Munger inversion to a plan before execution so hidden failure modes surface early
- `ww-swarm-planner` — Decide whether a goal should run as a single-agent task or as a multi-step swarm
- `ww-verifier` — Adversarially verify a worker's claimed output before Workweaver commits the step
- `zara-self-awareness` — Ground-truth description of Zara/tools/memory/guardrails/limits

### Business
- `ai-cto-ops` — Always-on autonomous CTO operator with governed authority, continuous monitoring, and executive memory
- `brainstorm` — Pre-planning pressure test and requirements capture
- `cold-outreach` — Plan and operate implemented cold-outreach slices; full W1-W5 automation remains internal/partial
- `document-processor` — Document-processing runtime capability and contract
- `research` — Codebase and web research
- `task-decomposition` — Decompose work into validated parallelizable slices
- `zoho-cli` — Zoho connector CLI truth and OAuth/scope debugging

### Dev
- `accessibility` — Accessibility review and WCAG guidance
- `aws-ops` — AWS/Terraform deployment, drift, cleanup, resilience
- `coding-standards` — Repo coding conventions
- `cross-review` — Independent second opinion for code, architecture, and decisions
- `docs-navigator` — Targeted navigation of canonical Workweaver docs
- `engineering-philosophy` — AI-native engineering tradeoffs and rewrite-vs-patch thinking
- `execution-discipline` — Fresh verification, review-response, and recovery discipline for agent work
- `frontend-artifacts` — Workweaver frontend implementation guidance for Mission-shell surfaces in the real stack
- `llm-council` — Multi-model consultation for stress testing, critical evaluation, blind spot detection, and creative ideation
- `local-docker-e2e` — Local Docker E2E verification
- `meta-skill` — Skill quality standard, templates, and validation checklist
- `performance` — Performance review and optimization
- `reuse-before-new` — Extreme reuse, consolidation, and legacy elimination before adding code
- `security` — Application security review
- `self-awareness` — Deep structural codebase analysis with AST-based census, validated proposals, and Codex adversarial review
- `skill-autoresearch` — Automated skill quality evaluation and improvement
- `tdd-implementation` — Reuse-aware preflight, RED-GREEN-REFACTOR implementation, and verification-first delivery
- `testing-suite` — TDD, regression tests, repro, smoke validation
- `uat-suite` — Canonical Workweaver UAT orchestrator at UAT/. Single-command exploratory user acceptance testing across REPLAY/SHADOW/LIVE modes with per-run sandbox tenants, shadow evidence ledger, strict assertions, silent-failure detection, LLM-as-judge for non-deterministic AI, and schema-driven UI/CLI/MCP/API parity.
- `ux-design` — Workweaver Mission-shell UX, progressive disclosure, and consistency guidance
- `zero-defect` — Structural defect handling, technical debt reduction, and no-zombie-test discipline

### Operator
- `ww-adversary` — Attack a plan to find exploits, weaknesses, and adversarial failure modes
- `ww-competitor-perspective` — Generate a plan fragment from a competitor intelligence perspective in a 360 role swarm
- `ww-convergence-agent` — Merge plan fragments from multiple role-agent perspectives into one coherent final plan
- `ww-customer-perspective` — Generate a plan fragment from the customer's perspective in a 360 role swarm
- `ww-defender` — Harden a plan against adversary findings, producing a more resilient version
- `ww-planner` — Generate a structured goal-to-DAG execution plan for a given business objective
- `ww-regulator-perspective` — Generate a plan fragment from a regulatory compliance perspective in a 360 role swarm
- `ww-revisor` — Revise a plan incorporating inverter failure-mode findings and verifier issues

## Inventory Facts
- Capability count: 41
- Business operator skills: ai-cto-ops, brainstorm, cold-outreach, document-processor, research, task-decomposition, ww-adversary, ww-competitor-perspective, ww-convergence-agent, ww-customer-perspective, ww-defender, ww-planner, ww-regulator-perspective, ww-revisor, zoho-cli
- Core runtime skills: voice-ai-integration, ww-inverter, ww-swarm-planner, ww-verifier, zara-self-awareness

## Public Boundaries
- Does NOT claim desktop, browser, or local device control.
- Pricing and deeper lifecycle detail are strongest inside authenticated product surfaces.
