eai-cli 0.0.0-bootstrap.0

package/resources/gofer/copilot-prompts/gofer_side.prompt.md

@@ -0,0 +1,58 @@
+---
+name: gofer:side
+description: Open a side conversation in the active CLI without disturbing the main pipeline state; resumable.
+agent: copilot-workspace
+tools:
+  - Read
+  - Grep
+  - Glob
+  - Bash
+  - WebSearch
+argument-hint: feature-name-or-description
+gofer:
+  workflowProfile: standard
+  canonicalSource: .specify/commands/gofer_side.md
+  canonicalChecksum: 53475e2a7dd37e5f9b87161efc109d27ff2551610cc2f16e6f2dbeeb46f47ef1
+  metadataSource: scripts/generate-commands.ts
+---
+
+
+# Gofer Side Conversation
+
+## Token And Cost Policy
+<!-- gofer:token-cost-policy:start -->
+
+Before spawning agents, calling tools, or loading large files:
+
+1. Treat `.specify/memory/gofer-model-policy.yaml` as the repo-owned source of truth for simple, medium, hard, and arbiter model routing. If it is missing, run `/gofer:bootstrap-workspace` before continuing.
+2. Use the cheapest capable model first.
+   - Claude: Haiku for scouting/extraction; Sonnet for normal implementation, synthesis, validation, and security; Opus for high-risk arbitration or release-critical failures.
+   - Codex/OpenAI: GPT mini for simple coding; GPT nano only for locate/classify/summarize/mechanical work; GPT-5.3-Codex or flagship GPT for tool-heavy coding, architecture, and release-critical validation.
+   - Gemini: Flash-Lite for cheap large-context scan/summarize; Flash for default research synthesis; Pro for large-context architecture or high-risk arbitration.
+   - Copilot: prefer Auto for simple and default work; ask the user before choosing a paid/high-tier picker model for hard security, architecture, or release gates.
+3. Keep raw tool output out of the main conversation context. Save stable findings to `.specify/specs/{feature}/context-bundle.md`, then work from summaries.
+4. Use provider prompt/context caching only for stable, non-secret prefixes: Gofer scaffold, AGENTS/CLAUDE/Copilot instructions, constitution, repo map, stage contracts, and validation rubric.
+5. Before continuing after large research, planning, implementation, or validation bursts, checkpoint the durable artifacts and compact/clear/resume context when the host supports it.
+6. Escalate model tier only when a cheaper pass is low-confidence, contradictory, security-sensitive, or blocking release quality.
+<!-- gofer:token-cost-policy:end -->
+
+Open a side conversation in the active CLI without disturbing the main Gofer
+pipeline state. Use this when you need to ask a quick clarifying question, run
+an exploratory search, or test an idea without polluting the working stage
+transcript or advancing the pipeline.
+
+Side conversations are scoped to the active session and are fully resumable:
+when you exit the side channel, the main pipeline context is restored exactly as
+it was, including the current stage, feature directory, and any pending
+auto-chain target. Memory writes inside the side channel are flagged so they are
+not auto-promoted into pipeline artifacts.
+
+Typical uses:
+
+- Asking "what does this function do?" mid-implement without losing tasks state
+- Comparing two design options before committing to one
+- Running an unrelated `/gofer:research` query against the same codebase
+
+To return to the main pipeline, exit the side channel; the next prompt resumes
+the prior stage. Side-channel transcripts are saved under
+`.specify/logs/side-channel/<timestamp>.jsonl` for later review.

package/resources/gofer/copilot-prompts/gofer_spec_summary.prompt.md

@@ -0,0 +1,125 @@
+---
+name: gofer:spec-summary
+description: Generate a business-friendly summary of feature value and scope.
+agent: copilot-workspace
+tools:
+  - Read
+  - Grep
+  - Glob
+  - Bash
+  - WebSearch
+argument-hint: feature-name-or-description
+gofer:
+  workflowProfile: standard
+  canonicalSource: .specify/commands/gofer_spec_summary.md
+  canonicalChecksum: c508bc3b357f7b6d919a514f630366651a5aafb07feabec81fbfa2066746e9cd
+  metadataSource: scripts/generate-commands.ts
+---
+
+## Workspace Preflight
+
+Before doing stage/helper work:
+
+1. Resolve the repository root.
+2. Check the core Gofer sentinels:
+   - `.specify/.gofer-version`
+   - `.specify/commands/0_gofer_start.md`
+   - `.specify/templates/spec-template.md`
+   - `.specify/templates/loop-contract-template.json`
+   - `.specify/templates/working-backwards-prfaq-template.md`
+   - `.specify/templates/business-owner-summary-template.md`
+   - `.specify/templates/cto-architecture-summary-template.md`
+   - `.specify/templates/ciso-security-summary-template.md`
+   - `.specify/templates/stakeholder-review-index-template.md`
+   - `.specify/scripts/bash/create-new-feature.sh`
+   - `.specify/scripts/node/parse-stage-command.mjs`
+   - `.specify/scripts/node/gofer-loop-audit.mjs`
+   - `.specify/scripts/hooks/post-tool-use.mjs`
+   - `.specify/scripts/powershell/install-optional-tools.ps1`
+   - `.specify/templates/gofer-model-policy.yaml`
+   - `.specify/memory/gofer-model-policy.yaml`
+   - `.specify/specs/`
+   - `.specify/memory/`
+3. Check host-specific repo-owned files when relevant:
+   - Claude: `AGENTS.md`, `CLAUDE.md`, `.claude/settings.json`
+   - Codex: `AGENTS.md`
+   - Copilot: `.github/copilot-instructions.md`
+   - VS Code extension mirrors Claude/Copilot/Gemini resources itself and should still keep the core scaffold healthy
+4. If the repo already has the workspace checker script, prefer running:
+   - `node .specify/scripts/node/gofer-workspace-check.mjs --host copilot --json`
+5. If the workspace is missing or stale, ask exactly:
+   - **"This repo is missing or stale for Gofer. Initialize/update it now?"**
+6. If the user says yes, run the Gofer workspace bootstrap helper and then resume this command from the top.
+7. If the user says no, stop and explain that Gofer stage/helper work depends on the repo-owned scaffold.
+
+
+# Gofer Spec Summary
+
+## EAI Platform Session Preflight
+
+Before any Gofer stage/helper command does pipeline work:
+
+1. Treat durable delivery as EAI Platform delivery by default, with Azure second
+   and every other stack only by explicit exception.
+2. Run `eai whoami` and confirm the EAI CLI is installed, the user is logged in,
+   and an active tenant is visible.
+3. If `eai` is missing, `eai whoami` fails, the token is expired, or no active
+   tenant is available, stop and run `/gofer:eai-first-run` or ask the user to
+   approve login/setup before continuing.
+4. For EAI app delivery, do not continue into research, specification, planning,
+   tasks, implementation, or validation until
+   `.specify/specs/{feature}/eai-preflight.md` records login, tenant, template,
+   app-readiness, and next-action evidence.
+5. Do not write tokens, secrets, private tenant IDs, or local `.env` values into
+   Gofer artifacts; record only product-safe readiness status and evidence.
+
+## Token And Cost Policy
+<!-- gofer:token-cost-policy:start -->
+
+Before spawning agents, calling tools, or loading large files:
+
+1. Treat `.specify/memory/gofer-model-policy.yaml` as the repo-owned source of truth for simple, medium, hard, and arbiter model routing. If it is missing, run `/gofer:bootstrap-workspace` before continuing.
+2. Use the cheapest capable model first.
+   - Claude: Haiku for scouting/extraction; Sonnet for normal implementation, synthesis, validation, and security; Opus for high-risk arbitration or release-critical failures.
+   - Codex/OpenAI: GPT mini for simple coding; GPT nano only for locate/classify/summarize/mechanical work; GPT-5.3-Codex or flagship GPT for tool-heavy coding, architecture, and release-critical validation.
+   - Gemini: Flash-Lite for cheap large-context scan/summarize; Flash for default research synthesis; Pro for large-context architecture or high-risk arbitration.
+   - Copilot: prefer Auto for simple and default work; ask the user before choosing a paid/high-tier picker model for hard security, architecture, or release gates.
+3. Keep raw tool output out of the main conversation context. Save stable findings to `.specify/specs/{feature}/context-bundle.md`, then work from summaries.
+4. Use provider prompt/context caching only for stable, non-secret prefixes: Gofer scaffold, AGENTS/CLAUDE/Copilot instructions, constitution, repo map, stage contracts, and validation rubric.
+5. Before continuing after large research, planning, implementation, or validation bursts, checkpoint the durable artifacts and compact/clear/resume context when the host supports it.
+6. Escalate model tier only when a cheaper pass is low-confidence, contradictory, security-sensitive, or blocking release quality.
+<!-- gofer:token-cost-policy:end -->
+
+Generate a business-friendly summary of the current feature and write it to
+`.specify/specs/{feature}/spec-summary.md`.
+
+Use this when a stakeholder or implementation team needs the plain-language
+purpose, expected outcomes, and scope boundaries without diving into the full
+spec.
+
+When you run this helper:
+
+1. Read the approved feature-local artifacts (`spec.md`, `plan.md`,
+   `contract-pack.md`, `quickstart.md`) and summarize only what is already in
+   scope.
+2. Keep the summary business-facing and humble. Do not turn it into a PRD or an
+   issue-tracker export.
+3. Write the artifact only to `.specify/specs/{feature}/spec-summary.md`.
+4. If the target file already exists, replace it and prepend a regeneration note
+   such as `<!-- regenerated at [ISO timestamp] -->`.
+5. Include the minimum provenance schema:
+   - `GeneratedAt`
+   - `SourceCommandId`
+   - `SourceInputs`
+   - `OverwriteNoticeWhenApplicable`
+
+The generated summary must contain these sections:
+
+- `## Provenance`
+- `## What`
+- `## Why`
+- `## Acceptance Criteria`
+- `## Out of Scope`
+
+Keep the content Gofer-owned. Do not copy upstream Matt Pocock skill text
+verbatim.

package/resources/gofer/copilot-prompts/gofer_tdd.prompt.md

@@ -0,0 +1,125 @@
+---
+name: gofer:tdd
+description: Guide a red-green-refactor loop tied to spec acceptance criteria.
+agent: copilot-workspace
+tools:
+  - Read
+  - Grep
+  - Glob
+  - Bash
+  - WebSearch
+argument-hint: feature-name-or-description
+gofer:
+  workflowProfile: standard
+  canonicalSource: .specify/commands/gofer_tdd.md
+  canonicalChecksum: 9f89107348de0a7d8eed69b2bcd2a6aa0652c51b1c678dcb3dc173116897c3d5
+  metadataSource: scripts/generate-commands.ts
+---
+
+## Workspace Preflight
+
+Before doing stage/helper work:
+
+1. Resolve the repository root.
+2. Check the core Gofer sentinels:
+   - `.specify/.gofer-version`
+   - `.specify/commands/0_gofer_start.md`
+   - `.specify/templates/spec-template.md`
+   - `.specify/templates/loop-contract-template.json`
+   - `.specify/templates/working-backwards-prfaq-template.md`
+   - `.specify/templates/business-owner-summary-template.md`
+   - `.specify/templates/cto-architecture-summary-template.md`
+   - `.specify/templates/ciso-security-summary-template.md`
+   - `.specify/templates/stakeholder-review-index-template.md`
+   - `.specify/scripts/bash/create-new-feature.sh`
+   - `.specify/scripts/node/parse-stage-command.mjs`
+   - `.specify/scripts/node/gofer-loop-audit.mjs`
+   - `.specify/scripts/hooks/post-tool-use.mjs`
+   - `.specify/scripts/powershell/install-optional-tools.ps1`
+   - `.specify/templates/gofer-model-policy.yaml`
+   - `.specify/memory/gofer-model-policy.yaml`
+   - `.specify/specs/`
+   - `.specify/memory/`
+3. Check host-specific repo-owned files when relevant:
+   - Claude: `AGENTS.md`, `CLAUDE.md`, `.claude/settings.json`
+   - Codex: `AGENTS.md`
+   - Copilot: `.github/copilot-instructions.md`
+   - VS Code extension mirrors Claude/Copilot/Gemini resources itself and should still keep the core scaffold healthy
+4. If the repo already has the workspace checker script, prefer running:
+   - `node .specify/scripts/node/gofer-workspace-check.mjs --host copilot --json`
+5. If the workspace is missing or stale, ask exactly:
+   - **"This repo is missing or stale for Gofer. Initialize/update it now?"**
+6. If the user says yes, run the Gofer workspace bootstrap helper and then resume this command from the top.
+7. If the user says no, stop and explain that Gofer stage/helper work depends on the repo-owned scaffold.
+
+
+# Gofer TDD
+
+## EAI Platform Session Preflight
+
+Before any Gofer stage/helper command does pipeline work:
+
+1. Treat durable delivery as EAI Platform delivery by default, with Azure second
+   and every other stack only by explicit exception.
+2. Run `eai whoami` and confirm the EAI CLI is installed, the user is logged in,
+   and an active tenant is visible.
+3. If `eai` is missing, `eai whoami` fails, the token is expired, or no active
+   tenant is available, stop and run `/gofer:eai-first-run` or ask the user to
+   approve login/setup before continuing.
+4. For EAI app delivery, do not continue into research, specification, planning,
+   tasks, implementation, or validation until
+   `.specify/specs/{feature}/eai-preflight.md` records login, tenant, template,
+   app-readiness, and next-action evidence.
+5. Do not write tokens, secrets, private tenant IDs, or local `.env` values into
+   Gofer artifacts; record only product-safe readiness status and evidence.
+
+## Token And Cost Policy
+<!-- gofer:token-cost-policy:start -->
+
+Before spawning agents, calling tools, or loading large files:
+
+1. Treat `.specify/memory/gofer-model-policy.yaml` as the repo-owned source of truth for simple, medium, hard, and arbiter model routing. If it is missing, run `/gofer:bootstrap-workspace` before continuing.
+2. Use the cheapest capable model first.
+   - Claude: Haiku for scouting/extraction; Sonnet for normal implementation, synthesis, validation, and security; Opus for high-risk arbitration or release-critical failures.
+   - Codex/OpenAI: GPT mini for simple coding; GPT nano only for locate/classify/summarize/mechanical work; GPT-5.3-Codex or flagship GPT for tool-heavy coding, architecture, and release-critical validation.
+   - Gemini: Flash-Lite for cheap large-context scan/summarize; Flash for default research synthesis; Pro for large-context architecture or high-risk arbitration.
+   - Copilot: prefer Auto for simple and default work; ask the user before choosing a paid/high-tier picker model for hard security, architecture, or release gates.
+3. Keep raw tool output out of the main conversation context. Save stable findings to `.specify/specs/{feature}/context-bundle.md`, then work from summaries.
+4. Use provider prompt/context caching only for stable, non-secret prefixes: Gofer scaffold, AGENTS/CLAUDE/Copilot instructions, constitution, repo map, stage contracts, and validation rubric.
+5. Before continuing after large research, planning, implementation, or validation bursts, checkpoint the durable artifacts and compact/clear/resume context when the host supports it.
+6. Escalate model tier only when a cheaper pass is low-confidence, contradictory, security-sensitive, or blocking release quality.
+<!-- gofer:token-cost-policy:end -->
+
+Guide a red-green-refactor loop for the active feature and write the cycle log
+to `.specify/specs/{feature}/tdd-session.md`.
+
+Use this helper when you want to work test-first inside the existing Gofer
+implementation flow without replacing `#5_gofer_implement` or `#9_gofer_tests`.
+
+When you run this helper:
+
+1. Read the feature acceptance criteria and the current task scope before
+   proposing test changes.
+2. Define the smallest failing test that proves the next slice of behavior is
+   missing.
+3. Keep the cycle explicit: red, green, then refactor.
+4. Record only feature-local artifacts and paths.
+5. Write the artifact only to `.specify/specs/{feature}/tdd-session.md`.
+6. If the target file already exists, replace it and prepend a regeneration note
+   such as `<!-- regenerated at [ISO timestamp] -->`.
+7. Include the minimum provenance schema:
+   - `GeneratedAt`
+   - `SourceCommandId`
+   - `SourceInputs`
+   - `OverwriteNoticeWhenApplicable`
+
+The generated TDD session must contain these sections:
+
+- `## Provenance`
+- `## Acceptance Criteria Linkage`
+- `## Red`
+- `## Green`
+- `## Refactor`
+
+Keep the content Gofer-owned. Do not copy upstream Matt Pocock skill text
+verbatim.

package/resources/gofer/copilot-prompts/gofer_vocabulary.prompt.md

@@ -0,0 +1,125 @@
+---
+name: gofer:vocabulary
+description: Extract domain terminology into a canonical feature glossary.
+agent: copilot-workspace
+tools:
+  - Read
+  - Grep
+  - Glob
+  - Bash
+  - WebSearch
+argument-hint: feature-name-or-description
+gofer:
+  workflowProfile: standard
+  canonicalSource: .specify/commands/gofer_vocabulary.md
+  canonicalChecksum: e6d7b30fdeba756916ca90099d67c888b6d0960daf6fcc526fa8c8bcec27bb28
+  metadataSource: scripts/generate-commands.ts
+---
+
+## Workspace Preflight
+
+Before doing stage/helper work:
+
+1. Resolve the repository root.
+2. Check the core Gofer sentinels:
+   - `.specify/.gofer-version`
+   - `.specify/commands/0_gofer_start.md`
+   - `.specify/templates/spec-template.md`
+   - `.specify/templates/loop-contract-template.json`
+   - `.specify/templates/working-backwards-prfaq-template.md`
+   - `.specify/templates/business-owner-summary-template.md`
+   - `.specify/templates/cto-architecture-summary-template.md`
+   - `.specify/templates/ciso-security-summary-template.md`
+   - `.specify/templates/stakeholder-review-index-template.md`
+   - `.specify/scripts/bash/create-new-feature.sh`
+   - `.specify/scripts/node/parse-stage-command.mjs`
+   - `.specify/scripts/node/gofer-loop-audit.mjs`
+   - `.specify/scripts/hooks/post-tool-use.mjs`
+   - `.specify/scripts/powershell/install-optional-tools.ps1`
+   - `.specify/templates/gofer-model-policy.yaml`
+   - `.specify/memory/gofer-model-policy.yaml`
+   - `.specify/specs/`
+   - `.specify/memory/`
+3. Check host-specific repo-owned files when relevant:
+   - Claude: `AGENTS.md`, `CLAUDE.md`, `.claude/settings.json`
+   - Codex: `AGENTS.md`
+   - Copilot: `.github/copilot-instructions.md`
+   - VS Code extension mirrors Claude/Copilot/Gemini resources itself and should still keep the core scaffold healthy
+4. If the repo already has the workspace checker script, prefer running:
+   - `node .specify/scripts/node/gofer-workspace-check.mjs --host copilot --json`
+5. If the workspace is missing or stale, ask exactly:
+   - **"This repo is missing or stale for Gofer. Initialize/update it now?"**
+6. If the user says yes, run the Gofer workspace bootstrap helper and then resume this command from the top.
+7. If the user says no, stop and explain that Gofer stage/helper work depends on the repo-owned scaffold.
+
+
+# Gofer Vocabulary
+
+## EAI Platform Session Preflight
+
+Before any Gofer stage/helper command does pipeline work:
+
+1. Treat durable delivery as EAI Platform delivery by default, with Azure second
+   and every other stack only by explicit exception.
+2. Run `eai whoami` and confirm the EAI CLI is installed, the user is logged in,
+   and an active tenant is visible.
+3. If `eai` is missing, `eai whoami` fails, the token is expired, or no active
+   tenant is available, stop and run `/gofer:eai-first-run` or ask the user to
+   approve login/setup before continuing.
+4. For EAI app delivery, do not continue into research, specification, planning,
+   tasks, implementation, or validation until
+   `.specify/specs/{feature}/eai-preflight.md` records login, tenant, template,
+   app-readiness, and next-action evidence.
+5. Do not write tokens, secrets, private tenant IDs, or local `.env` values into
+   Gofer artifacts; record only product-safe readiness status and evidence.
+
+## Token And Cost Policy
+<!-- gofer:token-cost-policy:start -->
+
+Before spawning agents, calling tools, or loading large files:
+
+1. Treat `.specify/memory/gofer-model-policy.yaml` as the repo-owned source of truth for simple, medium, hard, and arbiter model routing. If it is missing, run `/gofer:bootstrap-workspace` before continuing.
+2. Use the cheapest capable model first.
+   - Claude: Haiku for scouting/extraction; Sonnet for normal implementation, synthesis, validation, and security; Opus for high-risk arbitration or release-critical failures.
+   - Codex/OpenAI: GPT mini for simple coding; GPT nano only for locate/classify/summarize/mechanical work; GPT-5.3-Codex or flagship GPT for tool-heavy coding, architecture, and release-critical validation.
+   - Gemini: Flash-Lite for cheap large-context scan/summarize; Flash for default research synthesis; Pro for large-context architecture or high-risk arbitration.
+   - Copilot: prefer Auto for simple and default work; ask the user before choosing a paid/high-tier picker model for hard security, architecture, or release gates.
+3. Keep raw tool output out of the main conversation context. Save stable findings to `.specify/specs/{feature}/context-bundle.md`, then work from summaries.
+4. Use provider prompt/context caching only for stable, non-secret prefixes: Gofer scaffold, AGENTS/CLAUDE/Copilot instructions, constitution, repo map, stage contracts, and validation rubric.
+5. Before continuing after large research, planning, implementation, or validation bursts, checkpoint the durable artifacts and compact/clear/resume context when the host supports it.
+6. Escalate model tier only when a cheaper pass is low-confidence, contradictory, security-sensitive, or blocking release quality.
+<!-- gofer:token-cost-policy:end -->
+
+Extract the feature's shared domain language into a canonical glossary and
+write it to `.specify/specs/{feature}/glossary.md`.
+
+Use this when research, specification, contracts, or implementation rely on
+terms that need stable definitions across Claude, Copilot, Codex, and Gemini
+surfaces.
+
+When you run this helper:
+
+1. Read the feature-local context that already exists (`research.md`, `spec.md`,
+   `plan.md`, `contracts/`, `quickstart.md`) and ignore unrelated repository
+   content.
+2. Identify project-specific terms, acronyms, role names, workflow names, and
+   overloaded words that need precise definitions.
+3. Write the artifact only to `.specify/specs/{feature}/glossary.md`. Never
+   write to repo root or any provider-specific surface directory.
+4. If the target file already exists, replace it and prepend a regeneration note
+   such as `<!-- regenerated at [ISO timestamp] -->`.
+5. Include the minimum provenance schema:
+   - `GeneratedAt`
+   - `SourceCommandId`
+   - `SourceInputs`
+   - `OverwriteNoticeWhenApplicable`
+
+The generated glossary must contain these sections:
+
+- `## Provenance`
+- `## Term Entries`
+- `## Definitions`
+- `## Source Artifacts`
+
+Keep the content Gofer-owned and concise. Do not copy upstream Matt Pocock
+skill text verbatim.

package/resources/gofer/copilot-prompts/gofer_zoom_out.prompt.md

@@ -0,0 +1,125 @@
+---
+name: gofer:zoom-out
+description: Show how the current feature connects to broader system boundaries.
+agent: copilot-workspace
+tools:
+  - Read
+  - Grep
+  - Glob
+  - Bash
+  - WebSearch
+argument-hint: feature-name-or-description
+gofer:
+  workflowProfile: standard
+  canonicalSource: .specify/commands/gofer_zoom_out.md
+  canonicalChecksum: 8a56981788d24261081d7735fe31db912f58dad748e2116e1d776aa9f2d8b45c
+  metadataSource: scripts/generate-commands.ts
+---
+
+## Workspace Preflight
+
+Before doing stage/helper work:
+
+1. Resolve the repository root.
+2. Check the core Gofer sentinels:
+   - `.specify/.gofer-version`
+   - `.specify/commands/0_gofer_start.md`
+   - `.specify/templates/spec-template.md`
+   - `.specify/templates/loop-contract-template.json`
+   - `.specify/templates/working-backwards-prfaq-template.md`
+   - `.specify/templates/business-owner-summary-template.md`
+   - `.specify/templates/cto-architecture-summary-template.md`
+   - `.specify/templates/ciso-security-summary-template.md`
+   - `.specify/templates/stakeholder-review-index-template.md`
+   - `.specify/scripts/bash/create-new-feature.sh`
+   - `.specify/scripts/node/parse-stage-command.mjs`
+   - `.specify/scripts/node/gofer-loop-audit.mjs`
+   - `.specify/scripts/hooks/post-tool-use.mjs`
+   - `.specify/scripts/powershell/install-optional-tools.ps1`
+   - `.specify/templates/gofer-model-policy.yaml`
+   - `.specify/memory/gofer-model-policy.yaml`
+   - `.specify/specs/`
+   - `.specify/memory/`
+3. Check host-specific repo-owned files when relevant:
+   - Claude: `AGENTS.md`, `CLAUDE.md`, `.claude/settings.json`
+   - Codex: `AGENTS.md`
+   - Copilot: `.github/copilot-instructions.md`
+   - VS Code extension mirrors Claude/Copilot/Gemini resources itself and should still keep the core scaffold healthy
+4. If the repo already has the workspace checker script, prefer running:
+   - `node .specify/scripts/node/gofer-workspace-check.mjs --host copilot --json`
+5. If the workspace is missing or stale, ask exactly:
+   - **"This repo is missing or stale for Gofer. Initialize/update it now?"**
+6. If the user says yes, run the Gofer workspace bootstrap helper and then resume this command from the top.
+7. If the user says no, stop and explain that Gofer stage/helper work depends on the repo-owned scaffold.
+
+
+# Gofer Zoom Out
+
+## EAI Platform Session Preflight
+
+Before any Gofer stage/helper command does pipeline work:
+
+1. Treat durable delivery as EAI Platform delivery by default, with Azure second
+   and every other stack only by explicit exception.
+2. Run `eai whoami` and confirm the EAI CLI is installed, the user is logged in,
+   and an active tenant is visible.
+3. If `eai` is missing, `eai whoami` fails, the token is expired, or no active
+   tenant is available, stop and run `/gofer:eai-first-run` or ask the user to
+   approve login/setup before continuing.
+4. For EAI app delivery, do not continue into research, specification, planning,
+   tasks, implementation, or validation until
+   `.specify/specs/{feature}/eai-preflight.md` records login, tenant, template,
+   app-readiness, and next-action evidence.
+5. Do not write tokens, secrets, private tenant IDs, or local `.env` values into
+   Gofer artifacts; record only product-safe readiness status and evidence.
+
+## Token And Cost Policy
+<!-- gofer:token-cost-policy:start -->
+
+Before spawning agents, calling tools, or loading large files:
+
+1. Treat `.specify/memory/gofer-model-policy.yaml` as the repo-owned source of truth for simple, medium, hard, and arbiter model routing. If it is missing, run `/gofer:bootstrap-workspace` before continuing.
+2. Use the cheapest capable model first.
+   - Claude: Haiku for scouting/extraction; Sonnet for normal implementation, synthesis, validation, and security; Opus for high-risk arbitration or release-critical failures.
+   - Codex/OpenAI: GPT mini for simple coding; GPT nano only for locate/classify/summarize/mechanical work; GPT-5.3-Codex or flagship GPT for tool-heavy coding, architecture, and release-critical validation.
+   - Gemini: Flash-Lite for cheap large-context scan/summarize; Flash for default research synthesis; Pro for large-context architecture or high-risk arbitration.
+   - Copilot: prefer Auto for simple and default work; ask the user before choosing a paid/high-tier picker model for hard security, architecture, or release gates.
+3. Keep raw tool output out of the main conversation context. Save stable findings to `.specify/specs/{feature}/context-bundle.md`, then work from summaries.
+4. Use provider prompt/context caching only for stable, non-secret prefixes: Gofer scaffold, AGENTS/CLAUDE/Copilot instructions, constitution, repo map, stage contracts, and validation rubric.
+5. Before continuing after large research, planning, implementation, or validation bursts, checkpoint the durable artifacts and compact/clear/resume context when the host supports it.
+6. Escalate model tier only when a cheaper pass is low-confidence, contradictory, security-sensitive, or blocking release quality.
+<!-- gofer:token-cost-policy:end -->
+
+Expand the current feature into its surrounding system context and write the
+result to `.specify/specs/{feature}/zoom-out-report.md`.
+
+Use this helper when the feature needs broader architectural framing: upstream
+dependencies, downstream consumers, or cross-cutting risks that are easy to
+miss inside a narrow task view.
+
+When you run this helper:
+
+1. Read the approved feature-local artifacts and identify the current boundary
+   being changed.
+2. Map the upstream inputs, downstream consumers, and cross-cutting impacts that
+   materially affect the feature.
+3. Keep the output feature-local and architecture-focused.
+4. Write the artifact only to
+   `.specify/specs/{feature}/zoom-out-report.md`.
+5. If the target file already exists, replace it and prepend a regeneration note
+   such as `<!-- regenerated at [ISO timestamp] -->`.
+6. Include the minimum provenance schema:
+   - `GeneratedAt`
+   - `SourceCommandId`
+   - `SourceInputs`
+   - `OverwriteNoticeWhenApplicable`
+
+The generated zoom-out report must contain these sections:
+
+- `## Provenance`
+- `## Current Boundary`
+- `## Upstream/Downstream`
+- `## Cross-Cutting Impact`
+
+Keep the content Gofer-owned. Do not copy upstream Matt Pocock skill text
+verbatim.