Trust & safety
How the storefront treats safety, honesty, and compliance.
Documentation & policies
Documentation & policies
Every standing policy summarized in one place, plus a map of the documentation surfaces. Nothing here is buried behind a purchase; full policy documents ship with the catalog and inside every product kit.
Where to read next
Grouped by purpose. Each link targets a page in the production storefront. Status labels distinguish public policy, product guidance, and machine-readable reference material.
Agents & discovery
The machine-readable layer and how agents should read it.
Buying & commerce
Evaluate, compare, and use verified Polar checkout with explicit purchase authority.
Vault & subscriptions
Monthly and annual recurring Vault access through Polar.
Legal & standards
Licensing and the legal frame around every purchase.
The policies, summarized
License
Products and Vault downloads use a Single-Operator license: one operator plus the AI agents acting for that license holder. Broader Team, Agency, or Enterprise licenses are not currently offered. Licenses never transfer resale, redistribution, sublicense, or public-sharing rights to the raw kit files. Customer-created outputs remain theirs.
Refunds
14-day refund review if a product does not match its description. Missing, corrupted, or wrong-file delivery is corrected or reviewed for refund promptly. Duplicate purchases retain one entitlement and may be reviewed for a duplicate refund. Polar Vault renewals follow the published provider process; every x402 Vault renewal requires a new prepaid payment and never uses automatic wallet debit. Direct-wallet crypto outside x402 is inactive.
Onboarding
Every product ships a 10-minute quick start, full implementation docs, separate agent and human operating instructions, an integration pack (compatibility matrix, known conflicts, safe-install checklist, rollback guide, preview-before-apply), and free samples — so you can inspect real content before relying on anything.
Safety & install rules
Binding across every product: install into a clearly named new folder, never overwrite existing files, preview before applying, gate risky actions behind human approval, and ship a complete rollback path. These are catalog-level rules — see safety.
Support
Support scope and response targets are on the support page. The mailbox is live: support@signalharness.ai (verified 2026-07-07).
Frequently asked questions
What are these products?
Execution systems and upgrade packs for AI-assisted work: prompt harnesses, quality-control scorecards, agent safety gates, and complete business operating systems. Plain-text file kits (Markdown + JSON) that work with any AI assistant.
Will this break my existing setup?
No. Every product is non-destructive by default: it installs into one new, clearly named folder, never overwrites existing files, includes a preview-before-apply step, and ships with a rollback guide. Removing a product means deleting its folder.
Do I need coding skills?
No. Every product has a 10-15 minute manual quick-start path. Advanced integrations (MCP, automation platforms) are optional and documented separately.
Do these work with my AI tool?
The kits are plain text and tool-agnostic: Claude, ChatGPT, or any LLM assistant. Per-product compatibility matrices list tested integration patterns.
Will these make me money?
No product here promises income or results. They are systems designed to help you execute more consistently and delegate to AI more safely. Outcomes depend on your product, market, and execution.
Can my AI agent use the product on my behalf?
Yes -- each product includes AGENT-README and agent operating instructions. Your license covers agents acting for you. Risky actions remain gated behind your approval.