create-egregore 0.19.22 → 0.19.24

package/runtime/codex/.codex/skills/search/SKILL.md

@@ -1,79 +0,0 @@
----
-name: search
-description: Hybrid keyword + semantic recall over the org's shared memory (decisions, handoffs, quests, knowledge, meetings) when the user invokes /search or $search, or asks what the org knows/decided about a topic. Use proactively before answering questions about past work.
----
-
-# Egregore Search
-
-Native Codex Egregore skill. Fully local — no API calls, works in local and
-connected mode. Silence in your context does not mean silence in memory:
-search before answering from conversation context alone.
-
-## Flow
-
-1. Organizational recall must begin with the Egregore search entry point in
-   step 2. Do not resolve `memory/` to its sibling Git repository, change into
-   that absolute path, or improvise raw grep/listing as the first retrieval
-   action. Use direct file reads or grep only on evidence paths returned by the
-   ranked search.
-
-   Before any shell, file-read, grep, or other retrieval tool call, send exactly
-   one standalone commentary line:
-
-   - local/filesystem-only:
-     `⌕ Egregore · searching your organization’s memory`
-   - connected mode with automatic graph enrichment:
-     `⌕ Egregore Connect · searching your organization’s memory and relationships`
-
-   Send the line verbatim. Do not replace it with generic narration such as
-   “I’ll search org memory.” It must be visible before the tool call; captured
-   command output does not satisfy this requirement. Emit it once even when the
-   retrieval needs multiple queries or source reads.
-
-2. Run the search:
-
-```bash
-bash bin/search.sh query "your query" -n 6
-```
-
-- The command also prints the attribution as a direct-shell fallback. That
-  captured banner is diagnostic redundancy, not the user-visible product beat.
-  If the answer instead begins with a graph-only named read, follow the same
-  global Egregore Retrieval Beat before that query.
-- Default is hybrid (BM25 + vector + rerank); it falls back to keyword
-  search automatically when local models aren't built, with a note.
-- `--fast` — keyword only (BM25), instant. Use for exact terms: names,
-  filenames, error strings.
-- `--semantic` — vector only. Use when the user's wording likely differs
-  from the document's.
-- Graph enrichment is automatic in connected mode — hits carry `↳ graph`
-  state lines (status, recipient, artifact type, quest) with no flag.
-  `--no-enrich` skips it when speed matters; local mode never attempts it.
-
-3. Search the *concept*, not the user's literal sentence. If the first query
-   returns weak scores, run a second with different wording — one
-   keyword-ish, one conceptual.
-
-4. Results are snippets with `memory/...` file paths. Read the source file
-   for anything you will cite or act on — snippets truncate, and memory
-   files carry dates, authors, and caveats that change the answer.
-
-5. Answer directly, citing file paths. If results are empty or weak, say so
-   plainly — "nothing in memory on X" is a real answer. Do not present
-   conversation-context guesses as memory.
-
-## Routing
-
-- Searching *code* in the repo → use your normal file search, not this.
-- Relationship/status traversal in connected mode ("who claimed this
-  handoff?") → `bin/graph.sh`. The graph finds links between things;
-  search finds content inside things.
-
-## Maintenance
-
-- Fully self-maintaining: every query re-indexes incrementally and warms
-  missing embeddings in a background job. On a fresh machine the first
-  searches run in keyword mode while the semantic index builds itself
-  (~2GB one-time local model download); no manual setup.
-- `bash bin/search.sh reindex --embed` — force a full rebuild.
-- `bash bin/search.sh status` — index health.

package/runtime/codex/.codex/skills/session-view/SKILL.md

@@ -1,29 +0,0 @@
----
-name: session-view
-description: 'Turn the current working session into an expressive, self-contained HTML view of how it evolved — steady work as a quiet spine, decisions as forks showing the road taken beside the roads not taken. Use for "/session-view", "make sense of this session", "turn this session into an artifact", "show how this session evolved".'
----
-
-<!-- generated-by: bin/codex-sync-skills.sh -->
-
-# Egregore session-view Adapter
-
-This is fallback coverage for a long-tail Egregore workflow that has not been
-ported to a hand-written Codex-native skill yet.
-
-Use the project shell and filesystem directly. Do not invoke Claude Code
-commands. Translate interactive choices to structured Codex question tooling
-when it is available; otherwise render compact numbered choices with an
-`Other:` option and wait for the user.
-
-1. Read `.claude/skills/session-view/SKILL.md` for the workflow details.
-2. Run the referenced `bin/` scripts directly from Codex.
-3. Treat graph and publish steps as best-effort unless that workflow explicitly
-   says they are required.
-4. For every external notification, follow
-   `.claude/context/notification-consent.md`: plan without sending, then show
-   a separate exact Send / Edit / Cancel checkpoint. Never infer notification
-   consent from the workflow request or a batch approval.
-5. Keep local-mode behavior filesystem-first and avoid graph or notification
-   calls when `egregore.json` declares `"mode": "local"`.
-6. Never call the deprecated `egregore-handoff` CLI for Egregore project
-   handoffs.

package/runtime/codex/.codex/skills/setup/SKILL.md

@@ -1,29 +0,0 @@
----
-name: setup
-description: 'Fallback Codex adapter for the Egregore setup workflow.'
----
-
-<!-- generated-by: bin/codex-sync-skills.sh -->
-
-# Egregore setup Adapter
-
-This is fallback coverage for a long-tail Egregore workflow that has not been
-ported to a hand-written Codex-native skill yet.
-
-Use the project shell and filesystem directly. Do not invoke Claude Code
-commands. Translate interactive choices to structured Codex question tooling
-when it is available; otherwise render compact numbered choices with an
-`Other:` option and wait for the user.
-
-1. Read `.claude/skills/setup/SKILL.md` for the workflow details.
-2. Run the referenced `bin/` scripts directly from Codex.
-3. Treat graph and publish steps as best-effort unless that workflow explicitly
-   says they are required.
-4. For every external notification, follow
-   `.claude/context/notification-consent.md`: plan without sending, then show
-   a separate exact Send / Edit / Cancel checkpoint. Never infer notification
-   consent from the workflow request or a batch approval.
-5. Keep local-mode behavior filesystem-first and avoid graph or notification
-   calls when `egregore.json` declares `"mode": "local"`.
-6. Never call the deprecated `egregore-handoff` CLI for Egregore project
-   handoffs.

package/runtime/codex/.codex/skills/summon/SKILL.md

@@ -1,45 +0,0 @@
----
-name: summon
-description: 'Fallback Codex adapter for the Egregore summon workflow.'
----
-
-<!-- generated-by: bin/codex-sync-skills.sh -->
-
-# Egregore summon Adapter
-
-This is fallback coverage for a long-tail Egregore workflow that has not been
-ported to a hand-written Codex-native skill yet.
-
-Use the project shell and filesystem directly. Do not invoke Claude Code
-commands. Translate interactive choices to structured Codex question tooling
-when it is available; otherwise render compact numbered choices with an
-`Other:` option and wait for the user.
-
-## Structured UX parity
-
-This workflow has a Claude skill with user-visible structured output. After
-reading `.claude/skills/summon/SKILL.md`, reproduce the same visible UX in Codex:
-
-- Preserve TUI boxes, markdown tables, rich cards, browser artifact rendering,
-  exact confirmation blocks, and "no preamble" rules from the source skill.
-- Use the source skill's frame width, section order, labels, status footer,
-  and examples as the contract for the final response.
-- Never replace a required box/table/card/artifact view with a prose summary
-  unless the user explicitly asks for a summary.
-- When the source says to output a TUI box directly, paste that box as the
-  visible response, preferably in a `text` fenced block.
-- Never show raw JSON, raw command output, or unformatted script output when
-  the source skill requires formatted status or rendered output.
-
-1. Read `.claude/skills/summon/SKILL.md` for the workflow details.
-2. Run the referenced `bin/` scripts directly from Codex.
-3. Treat graph and publish steps as best-effort unless that workflow explicitly
-   says they are required.
-4. For every external notification, follow
-   `.claude/context/notification-consent.md`: plan without sending, then show
-   a separate exact Send / Edit / Cancel checkpoint. Never infer notification
-   consent from the workflow request or a batch approval.
-5. Keep local-mode behavior filesystem-first and avoid graph or notification
-   calls when `egregore.json` declares `"mode": "local"`.
-6. Never call the deprecated `egregore-handoff` CLI for Egregore project
-   handoffs.

package/runtime/codex/.codex/skills/task/SKILL.md

@@ -1,29 +0,0 @@
----
-name: task
-description: 'Create, import, inspect, approve, answer, retry, complete, or cancel work in the hosted Egregore Agent Desk. Use for `/task`, `$task`, “send this work to the agents,” “start these tasks,” Markdown development batches, or checking hosted task progress. This skill submits work to the hosted pipeline; it does not implement the task in the current local session.'
----
-
-<!-- generated-by: bin/codex-sync-skills.sh -->
-
-# Egregore task Adapter
-
-This is fallback coverage for a long-tail Egregore workflow that has not been
-ported to a hand-written Codex-native skill yet.
-
-Use the project shell and filesystem directly. Do not invoke Claude Code
-commands. Translate interactive choices to structured Codex question tooling
-when it is available; otherwise render compact numbered choices with an
-`Other:` option and wait for the user.
-
-1. Read `.claude/skills/task/SKILL.md` for the workflow details.
-2. Run the referenced `bin/` scripts directly from Codex.
-3. Treat graph and publish steps as best-effort unless that workflow explicitly
-   says they are required.
-4. For every external notification, follow
-   `.claude/context/notification-consent.md`: plan without sending, then show
-   a separate exact Send / Edit / Cancel checkpoint. Never infer notification
-   consent from the workflow request or a batch approval.
-5. Keep local-mode behavior filesystem-first and avoid graph or notification
-   calls when `egregore.json` declares `"mode": "local"`.
-6. Never call the deprecated `egregore-handoff` CLI for Egregore project
-   handoffs.

package/runtime/codex/.codex/skills/teams-connect/SKILL.md

@@ -1,29 +0,0 @@
----
-name: teams-connect
-description: 'Fallback Codex adapter for the Egregore teams-connect workflow.'
----
-
-<!-- generated-by: bin/codex-sync-skills.sh -->
-
-# Egregore teams-connect Adapter
-
-This is fallback coverage for a long-tail Egregore workflow that has not been
-ported to a hand-written Codex-native skill yet.
-
-Use the project shell and filesystem directly. Do not invoke Claude Code
-commands. Translate interactive choices to structured Codex question tooling
-when it is available; otherwise render compact numbered choices with an
-`Other:` option and wait for the user.
-
-1. Read `.claude/skills/teams-connect/SKILL.md` for the workflow details.
-2. Run the referenced `bin/` scripts directly from Codex.
-3. Treat graph and publish steps as best-effort unless that workflow explicitly
-   says they are required.
-4. For every external notification, follow
-   `.claude/context/notification-consent.md`: plan without sending, then show
-   a separate exact Send / Edit / Cancel checkpoint. Never infer notification
-   consent from the workflow request or a batch approval.
-5. Keep local-mode behavior filesystem-first and avoid graph or notification
-   calls when `egregore.json` declares `"mode": "local"`.
-6. Never call the deprecated `egregore-handoff` CLI for Egregore project
-   handoffs.

package/runtime/codex/.codex/skills/telegram-connect/SKILL.md

@@ -1,29 +0,0 @@
----
-name: telegram-connect
-description: 'Fallback Codex adapter for the Egregore telegram-connect workflow.'
----
-
-<!-- generated-by: bin/codex-sync-skills.sh -->
-
-# Egregore telegram-connect Adapter
-
-This is fallback coverage for a long-tail Egregore workflow that has not been
-ported to a hand-written Codex-native skill yet.
-
-Use the project shell and filesystem directly. Do not invoke Claude Code
-commands. Translate interactive choices to structured Codex question tooling
-when it is available; otherwise render compact numbered choices with an
-`Other:` option and wait for the user.
-
-1. Read `.claude/skills/telegram-connect/SKILL.md` for the workflow details.
-2. Run the referenced `bin/` scripts directly from Codex.
-3. Treat graph and publish steps as best-effort unless that workflow explicitly
-   says they are required.
-4. For every external notification, follow
-   `.claude/context/notification-consent.md`: plan without sending, then show
-   a separate exact Send / Edit / Cancel checkpoint. Never infer notification
-   consent from the workflow request or a batch approval.
-5. Keep local-mode behavior filesystem-first and avoid graph or notification
-   calls when `egregore.json` declares `"mode": "local"`.
-6. Never call the deprecated `egregore-handoff` CLI for Egregore project
-   handoffs.

package/runtime/codex/.codex/skills/the-spiral/SKILL.md

@@ -1,84 +0,0 @@
----
-name: the-spiral
-description: Run a Codex-native structured Socratic dialogue when the user invokes /the-spiral or $the-spiral to turn an intuition into rigorous, communicable output.
----
-
-# The Spiral
-
-Native Codex Egregore skill. The Spiral is a stateful Socratic process for
-developing, pressure-testing, and compressing a complex thesis.
-
-## Setup
-
-Gather three inputs, using structured Codex question tooling when available and
-plain one-at-a-time questions otherwise:
-
-1. Domain: what are we spiraling on?
-2. Purpose: what should the process produce?
-3. Known areas: what domains must be covered, if any?
-
-Create state files:
-
-```text
-spiral_state.json
-spiral_artifacts/ring1
-spiral_artifacts/ring2
-spiral_artifacts/ring3
-spiral_artifacts/ring4
-spiral_artifacts/ring5
-```
-
-State includes domain, purpose, addressable space, current ring, current loop,
-artifacts, emerged domains, parked threads, descents, and ring history.
-
-## Rings
-
-1. Seed: find the irreducible claim. Artifact:
-   `spiral_artifacts/ring1/seed.md`.
-2. Territory: map implications, tensions, contradictions, and open questions.
-   Artifact: `spiral_artifacts/ring2/territory.md`.
-3. Encounter: ground claims in reality. Research facts when needed and write
-   domain artifacts under `spiral_artifacts/ring3/`.
-4. Crucible: test objections and disconfirming evidence. Artifact:
-   `spiral_artifacts/ring4/crucible.md`.
-5. Compression: produce the final deliverables under
-   `spiral_artifacts/ring5/`.
-
-Ask one question at a time. After each alignment check, update
-`spiral_state.json`. Render a compact text progress view only at setup,
-alignment checks, ring transitions, descents, resume, and completion.
-
-## TUI Rendering
-
-Structured UX parity is required. The Spiral must present state as a structured
-terminal view at phase boundaries, not as loose paragraphs:
-
-- Show rings 1-5 with current/completed/parked status.
-- Show current thesis, emerged domains, parked threads, descents, and next
-  question or alignment check.
-- Render only at setup, alignment checks, ring transitions, descents, resume,
-  and completion.
-- On completion, render a final structured view with rings completed, loop
-  count, deliverable paths, emerged domains, and parked threads.
-- Do not render mid-flow while asking a Socratic question.
-
-## Flow Rules
-
-- If the user gives a marketing-like answer in Ring 1, push toward the real
-  conviction.
-- If an objection breaks an earlier assumption, offer a descent to the right
-  ring and record the reason.
-- Research is active only when factual claims need grounding or pressure.
-- Never force a domain that has not emerged from the user's answers.
-- If the user is tired or wants to pause, crystallize the current state and
-  stop at an alignment boundary.
-
-## Resume
-
-If `spiral_state.json` exists, summarize domain, ring, loop, and current
-artifacts, then ask whether to continue or start fresh.
-
-## Completion
-
-The process ends when Ring 5 deliverables satisfy the user. Report artifact
-paths, rings completed, loop count, emerged domains, and any parked threads.

package/runtime/codex/.codex/skills/thread/SKILL.md

@@ -1,29 +0,0 @@
----
-name: thread
-description: 'Threads — directed work-stream primitive: store the intent, project the state. Use for /thread new|view|dash|fold|review — creating a thread, viewing one, regenerating the heat-ring dashboard, folding a confirmed decision into stored intent, or running the iteration review. Also triggers when a #thread-<slug>-foldin:v1 block is pasted.'
----
-
-<!-- generated-by: bin/codex-sync-skills.sh -->
-
-# Egregore thread Adapter
-
-This is fallback coverage for a long-tail Egregore workflow that has not been
-ported to a hand-written Codex-native skill yet.
-
-Use the project shell and filesystem directly. Do not invoke Claude Code
-commands. Translate interactive choices to structured Codex question tooling
-when it is available; otherwise render compact numbered choices with an
-`Other:` option and wait for the user.
-
-1. Read `.claude/skills/thread/SKILL.md` for the workflow details.
-2. Run the referenced `bin/` scripts directly from Codex.
-3. Treat graph and publish steps as best-effort unless that workflow explicitly
-   says they are required.
-4. For every external notification, follow
-   `.claude/context/notification-consent.md`: plan without sending, then show
-   a separate exact Send / Edit / Cancel checkpoint. Never infer notification
-   consent from the workflow request or a batch approval.
-5. Keep local-mode behavior filesystem-first and avoid graph or notification
-   calls when `egregore.json` declares `"mode": "local"`.
-6. Never call the deprecated `egregore-handoff` CLI for Egregore project
-   handoffs.

package/runtime/codex/.codex/skills/todo/SKILL.md

@@ -1,45 +0,0 @@
----
-name: todo
-description: 'Fallback Codex adapter for the Egregore todo workflow.'
----
-
-<!-- generated-by: bin/codex-sync-skills.sh -->
-
-# Egregore todo Adapter
-
-This is fallback coverage for a long-tail Egregore workflow that has not been
-ported to a hand-written Codex-native skill yet.
-
-Use the project shell and filesystem directly. Do not invoke Claude Code
-commands. Translate interactive choices to structured Codex question tooling
-when it is available; otherwise render compact numbered choices with an
-`Other:` option and wait for the user.
-
-## Structured UX parity
-
-This workflow has a Claude skill with user-visible structured output. After
-reading `.claude/skills/todo/SKILL.md`, reproduce the same visible UX in Codex:
-
-- Preserve TUI boxes, markdown tables, rich cards, browser artifact rendering,
-  exact confirmation blocks, and "no preamble" rules from the source skill.
-- Use the source skill's frame width, section order, labels, status footer,
-  and examples as the contract for the final response.
-- Never replace a required box/table/card/artifact view with a prose summary
-  unless the user explicitly asks for a summary.
-- When the source says to output a TUI box directly, paste that box as the
-  visible response, preferably in a `text` fenced block.
-- Never show raw JSON, raw command output, or unformatted script output when
-  the source skill requires formatted status or rendered output.
-
-1. Read `.claude/skills/todo/SKILL.md` for the workflow details.
-2. Run the referenced `bin/` scripts directly from Codex.
-3. Treat graph and publish steps as best-effort unless that workflow explicitly
-   says they are required.
-4. For every external notification, follow
-   `.claude/context/notification-consent.md`: plan without sending, then show
-   a separate exact Send / Edit / Cancel checkpoint. Never infer notification
-   consent from the workflow request or a batch approval.
-5. Keep local-mode behavior filesystem-first and avoid graph or notification
-   calls when `egregore.json` declares `"mode": "local"`.
-6. Never call the deprecated `egregore-handoff` CLI for Egregore project
-   handoffs.

package/runtime/pi/.codex/skills/add/SKILL.md

@@ -1,29 +0,0 @@
----
-name: add
-description: 'Fallback Codex adapter for the Egregore add workflow.'
----
-
-<!-- generated-by: bin/codex-sync-skills.sh -->
-
-# Egregore add Adapter
-
-This is fallback coverage for a long-tail Egregore workflow that has not been
-ported to a hand-written Codex-native skill yet.
-
-Use the project shell and filesystem directly. Do not invoke Claude Code
-commands. Translate interactive choices to structured Codex question tooling
-when it is available; otherwise render compact numbered choices with an
-`Other:` option and wait for the user.
-
-1. Read `.claude/skills/add/SKILL.md` for the workflow details.
-2. Run the referenced `bin/` scripts directly from Codex.
-3. Treat graph and publish steps as best-effort unless that workflow explicitly
-   says they are required.
-4. For every external notification, follow
-   `.claude/context/notification-consent.md`: plan without sending, then show
-   a separate exact Send / Edit / Cancel checkpoint. Never infer notification
-   consent from the workflow request or a batch approval.
-5. Keep local-mode behavior filesystem-first and avoid graph or notification
-   calls when `egregore.json` declares `"mode": "local"`.
-6. Never call the deprecated `egregore-handoff` CLI for Egregore project
-   handoffs.

package/runtime/pi/.codex/skills/answer/SKILL.md

@@ -1,29 +0,0 @@
----
-name: answer
-description: 'Fallback Codex adapter for the Egregore answer workflow.'
----
-
-<!-- generated-by: bin/codex-sync-skills.sh -->
-
-# Egregore answer Adapter
-
-This is fallback coverage for a long-tail Egregore workflow that has not been
-ported to a hand-written Codex-native skill yet.
-
-Use the project shell and filesystem directly. Do not invoke Claude Code
-commands. Translate interactive choices to structured Codex question tooling
-when it is available; otherwise render compact numbered choices with an
-`Other:` option and wait for the user.
-
-1. Read `.claude/skills/answer/SKILL.md` for the workflow details.
-2. Run the referenced `bin/` scripts directly from Codex.
-3. Treat graph and publish steps as best-effort unless that workflow explicitly
-   says they are required.
-4. For every external notification, follow
-   `.claude/context/notification-consent.md`: plan without sending, then show
-   a separate exact Send / Edit / Cancel checkpoint. Never infer notification
-   consent from the workflow request or a batch approval.
-5. Keep local-mode behavior filesystem-first and avoid graph or notification
-   calls when `egregore.json` declares `"mode": "local"`.
-6. Never call the deprecated `egregore-handoff` CLI for Egregore project
-   handoffs.

package/runtime/pi/.codex/skills/ask/SKILL.md

@@ -1,67 +0,0 @@
----
-name: ask
-description: Ask an asynchronous Egregore question to a teammate when the user invokes /ask or $ask, or help the current user formulate inward reflection questions.
----
-
-# Egregore Ask
-
-Native Codex Egregore skill. Use this when the user wants to ask a teammate a
-question, run a topic by someone, or inspect pending questions for themselves.
-
-## Flow
-
-1. Run `bin/agent.sh sync`.
-2. Resolve the current author from `.egregore-state.json`, then git config.
-3. Load people from `memory/people/*.md`. Match target names against filename,
-   `name:`, `github:`, and heading text, case-insensitively. Strip leading `@`.
-4. If the user named a target, draft one to four concrete questions. In
-   connected mode, optionally gather context with `bin/graph.sh` queries, but
-   suppress raw JSON and continue from memory files if graph access fails.
-5. Preview the exact questions and target. Use structured Codex question
-   tooling when available; otherwise render:
-
-```text
-Save these questions?
-1. Save as-is
-2. Edit
-Other:
-```
-
-6. For each approved question, run:
-
-```bash
-bin/agent.sh ask --from "$AUTHOR" --to "$TARGET" --topic "$TOPIC" --question "$QUESTION"
-```
-
-Add harvest flags only when this ask is part of a harvest:
-`--harvest-id`, `--harvest-session-id`, `--turn`, `--question-intent`, and
-`--context-mode`.
-
-7. In connected mode, follow
-   `.claude/context/notification-consent.md`. Saving the questions is not
-   notification consent. Prepare without sending:
-
-```bash
-PLAN_JSON=$(bash bin/notify.sh plan send "$TARGET" "$AUTHOR has a question about \"$TOPIC\". Ask for \$activity to answer.")
-```
-
-Show the exact organization, recipient, channel, and message in a separate
-Send / Edit / Cancel checkpoint. Only a Send response to that preview permits
-one approve + dispatch.
-
-8. Report the created `memory/knowledge/questions/...` path and the truthful
-notification state.
-
-## Inward Mode
-
-If no target matches, do not create a question file. Generate a small set of
-questions for the user to answer in-session, grounded in current conversation
-and memory context. Offer to save the result as a reflection only after the user
-answers.
-
-## Rules
-
-- Person-targeted asks are asynchronous: store first; notification is a
-  separate exact-consent action.
-- Local mode is filesystem-only; skip graph and notification calls.
-- Do not use Claude Code commands.

package/runtime/pi/.codex/skills/audit/SKILL.md

@@ -1,29 +0,0 @@
----
-name: audit
-description: 'Fallback Codex adapter for the Egregore audit workflow.'
----
-
-<!-- generated-by: bin/codex-sync-skills.sh -->
-
-# Egregore audit Adapter
-
-This is fallback coverage for a long-tail Egregore workflow that has not been
-ported to a hand-written Codex-native skill yet.
-
-Use the project shell and filesystem directly. Do not invoke Claude Code
-commands. Translate interactive choices to structured Codex question tooling
-when it is available; otherwise render compact numbered choices with an
-`Other:` option and wait for the user.
-
-1. Read `.claude/skills/audit/SKILL.md` for the workflow details.
-2. Run the referenced `bin/` scripts directly from Codex.
-3. Treat graph and publish steps as best-effort unless that workflow explicitly
-   says they are required.
-4. For every external notification, follow
-   `.claude/context/notification-consent.md`: plan without sending, then show
-   a separate exact Send / Edit / Cancel checkpoint. Never infer notification
-   consent from the workflow request or a batch approval.
-5. Keep local-mode behavior filesystem-first and avoid graph or notification
-   calls when `egregore.json` declares `"mode": "local"`.
-6. Never call the deprecated `egregore-handoff` CLI for Egregore project
-   handoffs.

package/runtime/pi/.codex/skills/autosave/SKILL.md

@@ -1,29 +0,0 @@
----
-name: autosave
-description: 'View or change your personal autosave settings, and merge staged auto-save PRs. The consent surface for ambient session capture.'
----
-
-<!-- generated-by: bin/codex-sync-skills.sh -->
-
-# Egregore autosave Adapter
-
-This is fallback coverage for a long-tail Egregore workflow that has not been
-ported to a hand-written Codex-native skill yet.
-
-Use the project shell and filesystem directly. Do not invoke Claude Code
-commands. Translate interactive choices to structured Codex question tooling
-when it is available; otherwise render compact numbered choices with an
-`Other:` option and wait for the user.
-
-1. Read `.claude/skills/autosave/SKILL.md` for the workflow details.
-2. Run the referenced `bin/` scripts directly from Codex.
-3. Treat graph and publish steps as best-effort unless that workflow explicitly
-   says they are required.
-4. For every external notification, follow
-   `.claude/context/notification-consent.md`: plan without sending, then show
-   a separate exact Send / Edit / Cancel checkpoint. Never infer notification
-   consent from the workflow request or a batch approval.
-5. Keep local-mode behavior filesystem-first and avoid graph or notification
-   calls when `egregore.json` declares `"mode": "local"`.
-6. Never call the deprecated `egregore-handoff` CLI for Egregore project
-   handoffs.

package/runtime/pi/.codex/skills/branch/SKILL.md

@@ -1,29 +0,0 @@
----
-name: branch
-description: 'Fallback Codex adapter for the Egregore branch workflow.'
----
-
-<!-- generated-by: bin/codex-sync-skills.sh -->
-
-# Egregore branch Adapter
-
-This is fallback coverage for a long-tail Egregore workflow that has not been
-ported to a hand-written Codex-native skill yet.
-
-Use the project shell and filesystem directly. Do not invoke Claude Code
-commands. Translate interactive choices to structured Codex question tooling
-when it is available; otherwise render compact numbered choices with an
-`Other:` option and wait for the user.
-
-1. Read `.claude/skills/branch/SKILL.md` for the workflow details.
-2. Run the referenced `bin/` scripts directly from Codex.
-3. Treat graph and publish steps as best-effort unless that workflow explicitly
-   says they are required.
-4. For every external notification, follow
-   `.claude/context/notification-consent.md`: plan without sending, then show
-   a separate exact Send / Edit / Cancel checkpoint. Never infer notification
-   consent from the workflow request or a batch approval.
-5. Keep local-mode behavior filesystem-first and avoid graph or notification
-   calls when `egregore.json` declares `"mode": "local"`.
-6. Never call the deprecated `egregore-handoff` CLI for Egregore project
-   handoffs.