From eb9f9e63a9315bf4ef93271dc37d8d0c4d85a47e Mon Sep 17 00:00:00 2001 From: Moshu <1709219+DatMoshu@users.noreply.github.com> Date: Sun, 25 Oct 2026 09:50:12 -0500 Subject: [PATCH] docs: point agent instructions at the Org fleet core Adds the fleet-rules pointer and replaces the retired D:/_fable evidence tool and _agentShared rule paths with the Org core; regenerates every router. Co-Authored-By: Claude Opus 5.5 --- .cursor/rules/spritemotion.mdc | 6 +++--- .github/copilot-instructions.md | 6 +++--- AGENTS.md | 6 +++--- CLAUDE.md | 8 +++++--- 4 files changed, 14 insertions(+), 12 deletions(-) diff --git a/.cursor/rules/spritemotion.mdc b/.cursor/rules/spritemotion.mdc index 811a96d..6e44916 100644 --- a/.cursor/rules/spritemotion.mdc +++ b/.cursor/rules/spritemotion.mdc @@ -25,9 +25,9 @@ Cursor, Copilot and others). Before working: ## Evidence Save all evidence in `D:/_agentShared/runs///`; run-id is `__` (UTC, lower-case kebab). -Name files by `D:/_agentShared/rules/evidence_naming.md`; storage and retention follow `D:/_agentShared/rules/evidence_central.md`. -Start: `python D:/_fable/tools/evidence/run.py start --project PROJECT --seat SEAT --story STORY --topic TOPIC --repo REPO` (prints the folder). -Capture there or use `python D:/_fable/tools/evidence/run.py add FILE --seat SEAT --subject SUBJECT --kind KIND`; finish with `python D:/_fable/tools/evidence/run.py done --seat SEAT --summary RESULT` and put the run id in DONE. +Naming, storage and retention follow `D:/.projects/.Org/rules/done.md`. +Start: `python D:/.projects/.Org/tools/evidence/run.py start --project PROJECT --seat SEAT --story STORY --topic TOPIC --repo REPO` (prints the folder). +Capture there or use `python D:/.projects/.Org/tools/evidence/run.py add FILE --seat SEAT --subject SUBJECT --kind KIND`; finish with `python D:/.projects/.Org/tools/evidence/run.py done --seat SEAT --summary RESULT` and put the run id in DONE. Never save evidence in this repo/worktree; product source assets and test fixtures keep their own paths. Weekly purge removes only reviewed done runs older than 14 days; working/blocked/unreviewed runs are protected. ## Evidence emitter follow-up diff --git a/.github/copilot-instructions.md b/.github/copilot-instructions.md index 4cc744b..4cd6e8e 100644 --- a/.github/copilot-instructions.md +++ b/.github/copilot-instructions.md @@ -20,9 +20,9 @@ Cursor, Copilot and others). Before working: ## Evidence Save all evidence in `D:/_agentShared/runs///`; run-id is `__` (UTC, lower-case kebab). -Name files by `D:/_agentShared/rules/evidence_naming.md`; storage and retention follow `D:/_agentShared/rules/evidence_central.md`. -Start: `python D:/_fable/tools/evidence/run.py start --project PROJECT --seat SEAT --story STORY --topic TOPIC --repo REPO` (prints the folder). -Capture there or use `python D:/_fable/tools/evidence/run.py add FILE --seat SEAT --subject SUBJECT --kind KIND`; finish with `python D:/_fable/tools/evidence/run.py done --seat SEAT --summary RESULT` and put the run id in DONE. +Naming, storage and retention follow `D:/.projects/.Org/rules/done.md`. +Start: `python D:/.projects/.Org/tools/evidence/run.py start --project PROJECT --seat SEAT --story STORY --topic TOPIC --repo REPO` (prints the folder). +Capture there or use `python D:/.projects/.Org/tools/evidence/run.py add FILE --seat SEAT --subject SUBJECT --kind KIND`; finish with `python D:/.projects/.Org/tools/evidence/run.py done --seat SEAT --summary RESULT` and put the run id in DONE. Never save evidence in this repo/worktree; product source assets and test fixtures keep their own paths. Weekly purge removes only reviewed done runs older than 14 days; working/blocked/unreviewed runs are protected. ## Evidence emitter follow-up diff --git a/AGENTS.md b/AGENTS.md index 4cc744b..4cd6e8e 100644 --- a/AGENTS.md +++ b/AGENTS.md @@ -20,9 +20,9 @@ Cursor, Copilot and others). Before working: ## Evidence Save all evidence in `D:/_agentShared/runs///`; run-id is `__` (UTC, lower-case kebab). -Name files by `D:/_agentShared/rules/evidence_naming.md`; storage and retention follow `D:/_agentShared/rules/evidence_central.md`. -Start: `python D:/_fable/tools/evidence/run.py start --project PROJECT --seat SEAT --story STORY --topic TOPIC --repo REPO` (prints the folder). -Capture there or use `python D:/_fable/tools/evidence/run.py add FILE --seat SEAT --subject SUBJECT --kind KIND`; finish with `python D:/_fable/tools/evidence/run.py done --seat SEAT --summary RESULT` and put the run id in DONE. +Naming, storage and retention follow `D:/.projects/.Org/rules/done.md`. +Start: `python D:/.projects/.Org/tools/evidence/run.py start --project PROJECT --seat SEAT --story STORY --topic TOPIC --repo REPO` (prints the folder). +Capture there or use `python D:/.projects/.Org/tools/evidence/run.py add FILE --seat SEAT --subject SUBJECT --kind KIND`; finish with `python D:/.projects/.Org/tools/evidence/run.py done --seat SEAT --summary RESULT` and put the run id in DONE. Never save evidence in this repo/worktree; product source assets and test fixtures keep their own paths. Weekly purge removes only reviewed done runs older than 14 days; working/blocked/unreviewed runs are protected. ## Evidence emitter follow-up diff --git a/CLAUDE.md b/CLAUDE.md index 324befd..0601303 100644 --- a/CLAUDE.md +++ b/CLAUDE.md @@ -1,5 +1,7 @@ # SpriteMotion: project rules for coding agents +Fleet rules: read D:/.projects/.Org/rules/00_index.md first (roles in D:/.projects/.Org/roles/). Seats are `chief`, `-lead` and `-worker-`, launched by `D:/.projects/.Org/launchers/{chief,lead,worker}.bat`; mail is `org mail`, leases `org lease`, state `org seat`, budget `org budget check` (`org` = `python D:/.projects/.Org/tools/org/run.py`). The old fleet_inbox, switchboard, roster and exec-fable/exec-astra names are retired (2026-10-24). + Shared by every agent (Claude Code, Codex, Cursor, Copilot). Workflows live in `.claude/skills/`; the routers (`AGENTS.md`, `.github/copilot-instructions.md`, `.cursor/rules/`) are generated from them by `python tools/agents/run.py`. Edit this file or a skill, never a router. @@ -56,9 +58,9 @@ headless Blender run, not only a syntax check. ## Evidence Save all evidence in `D:/_agentShared/runs///`; run-id is `__` (UTC, lower-case kebab). -Name files by `D:/_agentShared/rules/evidence_naming.md`; storage and retention follow `D:/_agentShared/rules/evidence_central.md`. -Start: `python D:/_fable/tools/evidence/run.py start --project PROJECT --seat SEAT --story STORY --topic TOPIC --repo REPO` (prints the folder). -Capture there or use `python D:/_fable/tools/evidence/run.py add FILE --seat SEAT --subject SUBJECT --kind KIND`; finish with `python D:/_fable/tools/evidence/run.py done --seat SEAT --summary RESULT` and put the run id in DONE. +Naming, storage and retention follow `D:/.projects/.Org/rules/done.md`. +Start: `python D:/.projects/.Org/tools/evidence/run.py start --project PROJECT --seat SEAT --story STORY --topic TOPIC --repo REPO` (prints the folder). +Capture there or use `python D:/.projects/.Org/tools/evidence/run.py add FILE --seat SEAT --subject SUBJECT --kind KIND`; finish with `python D:/.projects/.Org/tools/evidence/run.py done --seat SEAT --summary RESULT` and put the run id in DONE. Never save evidence in this repo/worktree; product source assets and test fixtures keep their own paths. Weekly purge removes only reviewed done runs older than 14 days; working/blocked/unreviewed runs are protected. ## Evidence emitter follow-up