The register
Standards: 11
In force: 1
Draft: 10
standards.thefocus.ai
Guides: 12
Research notes: 6
Decision records: 4
Prompts: 1
Skills: 14
The register
Every document The Focus AI has issued, with what it binds and whether it is
in force. A standard in Draft binds nothing: it reflects current
understanding and is awaiting ratification, which only a person can give
(STD-001 §3.14). Nothing below the standards binds anything at
all.
This page is generated from the documents themselves. A standard that does
not conform to STD-001, or a document that does not conform to
STD-006, does not render, and so does not appear here.
Standards
| Ref | Title | Category | Applies to | Status | Clauses |
|---|---|---|---|---|---|
| STD-001 | How a standard is written | Process | Every document in standards/ | In force | 18 |
| STD-002 | Issues and agent workflow | Practice | Every repository worked by an agent | Draft | 19 |
| STD-003 | Pull requests and branches | Practice | Every repository worked by an agent | Draft | 21 |
| STD-004 | Repository tooling and tasks | Practice | Every repository | Draft | 10 |
| STD-005 | Agent environment bootstrap | Practice | Every repository worked by a cloud agent | Draft | 9 |
| STD-006 | Documents and provenance | Process | Every document in this repository | Draft | 12 |
| STD-007 | Secrets and credentials | Practice | Every repository | Draft | 13 |
| STD-008 | Deployment | Practice | Every repository that deploys | Draft | 14 |
| STD-009 | Authentication | Practice | Every client-facing project with sign-in, and every product acting as an OAuth client | Draft | 12 |
| STD-010 | Agent services | Practice | A2A-discoverable agent services | Draft | 7 |
| STD-011 | Agent skills | Practice | Every repository worked by an agent | Draft | 7 |
Guides
Guides carry explanation, walkthroughs and reference implementations. They are not standards and bind nothing (STD-001 §4).
| Ref | Title | Relates to | Status |
|---|---|---|---|
| GDE-001 | Building an A2A agent | STD-007, STD-008, STD-009, STD-010 | Current |
| GDE-002 | Clerk authentication setup | STD-009 | Current |
| GDE-003 | Project secrets with fnox and 1Password | STD-007 | Current |
| GDE-004 | GCE Gaia runtime | STD-008 | Current |
| GDE-005 | GCP deployment | STD-008 | Current |
| GDE-006 | Local development environment | STD-004, STD-005, STD-007, STD-008 | Current |
| GDE-007 | pi extensions | STD-005 | Current |
| GDE-008 | Standard skill set | STD-002, STD-011 | Current |
| GDE-009 | Technology defaults | STD-004, STD-007, STD-009, STD-011 | Current |
| GDE-010 | Vercel deployment | STD-008 | Current |
| GDE-011 | Bootstrapping a new project | STD-001, STD-004, STD-005, STD-007, STD-008, STD-011 | Current |
| GDE-012 | Auditing an existing project against the standards | STD-001, STD-002, STD-003, STD-004, STD-005, STD-006, STD-007, STD-008, STD-011 | Current |
Research notes
Notes record what was true when somebody checked, against sources they cite. They are not standards and bind nothing (STD-001 §4).
| Ref | Title | Relates to | Verified | Status |
|---|---|---|---|---|
| NOTE-001 | gcloud Cloud Run Deployments: Multi-Project Configuration Guide | — | 2026-02-02 | Current |
| NOTE-002 | skills CLI: Technical Implementation Guide for Agent Skill Management | — | 2026-06-25 | Current |
| NOTE-003 | Creative typography: research notes for DESIGN.md work | — | 2026-07-23 | Current |
| NOTE-004 | best-practices/ inventory: what migrates, what stays, what is missing | STD-001, STD-003, STD-004, STD-005, STD-006, STD-007, STD-011 | 2026-07-26 | Current |
| NOTE-005 | Standards gap register: what is practised but not written | STD-001, STD-002, STD-003, STD-004, STD-005, STD-006, STD-007 | 2026-07-26 | Current |
| NOTE-006 | Migration coverage map: every source rule to its clause | STD-004, STD-006, STD-007, STD-008, STD-009, STD-010, STD-011 | 2026-07-26 | Current |
Decision records
Decision records settle a question for this repository and say what would reopen it. They are not standards and bind nothing (STD-001 §4).
| Ref | Title | Relates to | Status |
|---|---|---|---|
| ADR-001 | Standards Dockerfile redesign | — | Accepted |
| ADR-002 | Wrapping a habitat in an Electron app | — | Proposed |
| ADR-003 | standards.thefocus.ai design decision log | — | Accepted |
| ADR-004 | Ticket selection is judged, not sorted | STD-002, STD-006 | Accepted |
Prompts
Prompts are instructions an agent runs. They are input to a tool, not guidance for a person. They are not standards and bind nothing (STD-001 §4).
| Ref | Title | Relates to | Status |
|---|---|---|---|
| PRM-001 | Generate the standards Dockerfile | — | Current |
Skills
Capabilities an agent loads when its triggers match, installable with
skills add. A skill is cited by name and numbered here;
STD-011 governs how one is installed and pinned.
| Ref | Skill | Provenance | Relates to |
|---|---|---|---|
| SKL-001 | activity-report | Authored here | STD-007 |
| SKL-002 | browser-automation | Vendored | — |
| SKL-003 | develop-design-md | Authored here | — |
| SKL-004 | focus-ai-brand | Vendored | — |
| SKL-005 | nano-banana-imagegen | Vendored | — |
| SKL-006 | nano-banana-videogen | Vendored | — |
| SKL-007 | paged-reports | Authored here | — |
| SKL-008 | pptx-decks | Authored here | — |
| SKL-009 | research-methodology | Vendored | — |
| SKL-010 | setup-project | Authored here | STD-004, STD-007, STD-011 |
| SKL-011 | standardize-project | Authored here | STD-007, STD-008 |
| SKL-012 | tech-research | Authored here | — |
| SKL-013 | work-next-issue | Authored here | STD-002, STD-003 |
| SKL-014 | standards-map | Authored here | STD-007, STD-008 |