rodo-spec 0.0.0-stage → 2.2.0

package/dist/kit/.agents/skills/rodo/templates/spec.md

@@ -0,0 +1,55 @@
+# {{NNN}} — {{title}}: technical spec
+
+> Source of truth for the coder. Written in the technical language (`project.json`). Anything not here is not expected.
+> Every rule has a bold ID (**R1**, **R2** ...). If the card has scenarios, tag each scenario with the rules it proves (`@R1`) and try to cover every rule.
+
+## 1. Summary
+
+{{three lines}}
+
+## 2. Rules
+
+- **R1** {{rule}}
+- **R2** {{rule, including the exact error message the user sees}}
+
+## 3. Data
+
+{{new tables/fields, types, constraints}}
+
+## 4. API / logic
+
+{{endpoints, inputs, outputs, errors}}
+
+## 5. Screens
+
+{{what is shown where; exact message texts}}
+
+## 6. Files to create or change
+
+- `{{path}}` — {{what}}
+
+## 7. Edge cases
+
+- {{case}}
+
+## 8. Acceptance run (if the card has scenarios)
+
+Optional. `rodo test` (the card's `test.command`, or `testCommand` from `project.json` with `{NNN}` = `{{NNN}}`).
+
+## Acceptance criteria
+
+- **Must happen:** {{observable results; link rules, e.g. R1, R2}}
+- **Must not happen:** {{regressions or side effects that would be wrong}}
+- **Done when:** {{how we know: which checks pass, what the user can see}}
+
+## Documentation Impact
+
+- {{path}} — {{what changes}}   <!-- or: - none — why -->
+
+## Open Questions
+
+- {{question}} — default if no answer: {{default}}   <!-- or: - none -->
+
+## Decisions
+
+- {{decision}}. Because {{reason}}.

package/dist/kit/.agents/skills/rodo/testing.md

@@ -0,0 +1,45 @@
+> یادداشت فارسی: تست اختیاری است ولی پیشنهاد می‌شود، به‌خصوص برای قابلیت‌ها. این فایل قواعد نوشتن تست e2e (مرورگری) و چک‌لیست قبل از دیپلوی است.
+
+# Testing and deploy playbook
+
+Tests are optional in the RODO method, but recommended, especially for features. Nothing blocks work when there are none. Read this when you write or fix end-to-end (browser/API) tests, step definitions or CI test jobs, or prepare a deploy. It adds to your role's playbook.
+
+## Cycle scenarios (playwright-bdd)
+
+- Scenarios live in `rodo/<project>/cycles/NNN-slug/acceptance.feature`, tagged `@cycle-NNN`; each scenario (or its `Rule:`) is tagged with the spec rule IDs it proves (`@R1`). Try to give every `**Rn**` rule in spec.md at least one scenario.
+- Step definitions live in the `stepsDirs` folders of `project.json` and must really drive the app and really assert.
+- Run one card's tests with `rodo test`: it runs the card's `test.command` (or `testCommand` from `project.json`, replacing `{NNN}`) and records the failure signature. v2 supports the `web` (Playwright + optional playwright-bdd) and `api` (Jest/HTTP) adapters; other adapters work as plain commands with no special parsing. Tick `تست پذیرش پاس شد` only after a real pass. Note: `bddgen` needs `--tags @cycle-NNN` because it fails for the whole project if any cycle still lacks step definitions.
+- Cancelled cycles: add the tag `@cancelled` and exclude it in the Playwright config (`tags: 'not @cancelled'`).
+- **Never weaken, delete or skip a scenario or test to make code pass.**
+
+## E2E test conventions
+
+1. **Set up through the API, test through the UI.** Create users, workspaces and records with small API helpers; use the browser only for the flow under test. Faster and less flaky.
+2. **Isolated, unique data.** Every test creates its own data with a unique name built from a run prefix (`E2E_PREFIX`, default a timestamp). Never depend on data from another test or on test order.
+3. **No fixed sleeps.** No `waitForTimeout`/`sleep`. Wait for the result itself (web-first assertions, `expect(...).toPass()` for retries). Raise a timeout only for a known slow step and say why.
+4. **Stable selectors.** Prefer `data-testid` and roles. If you add a test id to app code, keep it in the same commit as the test.
+5. **Cover failure and permission paths**, not only happy paths: validation errors, 401 for anonymous, 403/404 for strangers and lower roles, removed members, expired sessions, two users at once (realtime / conflict) where the feature promises it.
+6. **Flake check before you call it done:** run the suite twice in parallel (`--repeat-each=2 --workers=4`). A test that fails once is a bug in the test or the app; find which. Never "fix" it with a retry or a sleep alone.
+7. **`fixme` only with a reason.** Allowed only in the project's own regression suite (never in a cycle's `acceptance.feature`), only when the expected behaviour is a product decision nobody has made. Write the open question in a comment next to it and list it in your report.
+8. **Typecheck the tests.** Give the e2e folder its own `tsconfig.json` and run it in CI, especially when the app build ignores type errors.
+
+## Running against a real (shared) server
+
+- One switch, e.g. `E2E_BASE_URL=https://...`: same suite, no local web servers, `workers=1` (shared rate limits), no retries.
+- **Throwaway data only:** unique prefix, never touch existing data, no deploys or restarts from tests.
+- Record every user/workspace the run creates in a git-ignored file and delete what the API allows in a global teardown. Report what could not be deleted.
+- If the server is known to crash on some route (e.g. an image optimizer on an old CPU), block that route in remote mode instead of crashing a shared server.
+
+## CI
+
+- Run the e2e suite in CI with a real database service (e.g. Postgres container), migrations applied, backend + frontend started by the test runner.
+- Upload the HTML report / traces as an artifact when it fails.
+- A good CI runs lint, typecheck (incl. tests), unit tests, API tests and e2e. If CI is red, say so in the card's Progress notes before you call the card done.
+
+## Deploy checklist (things that bit us)
+
+- **Old CPUs and native modules.** Prebuilt native packages (sharp/libvips, some crypto, ML runtimes) need newer CPU instructions (SSE4.2/AVX). On older CPUs the process dies with **exit code 132 (SIGILL)**, which looks like a random crash loop. Check the target: `grep -o -w 'sse4_2\|avx2' /proc/cpuinfo | sort -u`. Make such features opt-in, or strip the module from the image. Reproduce locally with `qemu-x86_64 -cpu phenom node ...`.
+- **Reverse proxy.** Behind Nginx/another proxy, configure the app to trust the proxy (`trust proxy`), or every per-IP rate limit becomes one global limit.
+- **Session tokens.** Rotating refresh tokens need a single-flight refresh on the client and a short reuse grace window on the server, or parallel tabs log users out.
+- **Migrations** run once per deploy, before the new code serves traffic; a deploy that adds a column needs the migration in the same release.
+- After deploy: health endpoint, then the e2e suite in remote mode.

package/dist/kit/.gitattributes

@@ -0,0 +1 @@
+* text eol=lf

package/dist/kit/KIT_VERSION

@@ -0,0 +1 @@
+2.2.0

package/dist/kit/MIGRATION.md

@@ -0,0 +1,71 @@
+# RODO kit migration notes
+
+Kit 2.1+ is installed and updated by `rodo init` (the 2.0 scripts `rodo-install.cjs` and `rodo-update.cjs` are replaced by it). Do the steps for every version you skipped.
+
+## 2.1.0 → 2.2.0 (English templates, rodo new, safer sync)
+
+What changes:
+- **English.** New projects choose the plain-language docs language at `rodo init` (`--lang en|fa`, default from the system locale). English projects get `project-template-en/` and the skill templates in `templates/en/`.
+- **`rodo new "<title>"`** creates the next cycle folder with the card, `simple.md` and `spec.md`.
+- **Card titles** no longer start with the number (the number comes from the folder). Old titles keep working; status and resume no longer print the number twice.
+- **MAP checks** accept relative links and English headings (`## Done`, `## Cancelled`); a done cycle missing from MAP gets `CYCLE_DONE_NOT_IN_MAP`.
+- **Cards and Git.** Without a board, card files are committed (the `.gitignore` card line is added only when a board is bound).
+- **Sync.** Progress notes survive a board edit that drops them; when the board wins a conflict your file is saved in `.rodo/conflicts/`.
+
+Steps: run `rodo init` (unchanged kit files update by themselves; files you changed get a `.rodo-new` copy). `RULES.md` is never replaced: compare it with the kit's new one (new "Quick start" section). Commit: `git commit -am "chore: RODO kit 2.2.0"`.
+
+## 2.0.0 → 2.1.0 (several projects per repo, session protocol, adaptive workflow)
+
+What changes:
+- **Layout.** `rodo/MAP.md`, `MEMORY.md`, `RULES.md`, `CHANGELOG.md`, `project/`, `cycles/` and `config.json` move into `rodo/<project>/` (`config.json` becomes `project.json`; `.kit-version` becomes `rodo/kit.json`). One repo can now hold several RODO projects, each bound to one board.
+- **Session protocol.** The AI looks at git without pulling, runs `rodo sync` and `rodo resume` at the start, and `rodo sync` at the end.
+- **Statuses.** The reason for `blocked` / `cancelled` now goes in the `statusReason` key of the card. Old `- Blocker:` lines still work, and `rodo sync` adds the key for you. Blocked cards also get a `blockerType`.
+- **Adaptive workflow.** Three levels (QUICK, STANDARD, MAJOR), at most 5 questions before planning, `design.md` for MAJOR only, and a documentation consistency check in VERIFY. New optional card key `workflow` (local only).
+- **New card parts.** `scope`, `test`, and `## Decision:` sections (answered on the web).
+
+Steps:
+
+1. Install the CLI, then in the repo run `rodo init --method-only --project <slug>` (add `--dry-run` first to see the plan). It moves the files, updates unchanged kit files, keeps files you changed as `<file>.rodo-new`, and never touches `scripts/rodo-*.cjs`.
+2. Compare each `.rodo-new` file with yours and merge by hand. `RULES.md` is never replaced automatically: compare it with the kit's new `RULES.md` (new sections: session start/end, stuck, decisions, scope, three work sizes, only important questions, docs always match the code).
+3. Connect the board: run `rodo init` in your own terminal and type the board name when asked. The AI never creates boards.
+4. Old `scripts/rodo-install.cjs`, `rodo-update.cjs` and `scripts/lib/core.cjs` are not needed any more. `rodo doctor` lists them as INFO; delete them when you are ready.
+5. Commit: `git add -A && git commit -m "chore: RODO kit 2.1.0"`.
+
+
+## 1.x → 2.0.0 (method only, no enforcement)
+
+2.0.0 removes every check that could block a commit or stop work. RODO is now a method the AI follows. Card-format checks move to the RODO watcher / CLI (`rodo validate`), which only reports warnings and errors.
+
+What is gone: `.githooks/` (pre-commit, commit-msg), `.github/workflows/rodo.yml`, `.claude/settings.json` + `.claude/hooks/rodo-guard.cjs`, `scripts/rodo-lint.cjs`, `scripts/rodo-lint`, `scripts/rodo-done.cjs`, `scripts/rodo-approval.cjs`, `acceptance-result.json` files, the approval fingerprint `[approval: hash]`, `x-failures`.
+
+Manual steps (in your project folder, in a terminal):
+
+1. **Turn off the old git hooks:**
+   ```
+   git config --unset core.hooksPath
+   ```
+   (If it says nothing or "not found", hooks were not on. Fine.)
+2. **Delete the old enforcement files** (check first that you did not add your own things to them):
+   - Windows PowerShell:
+     ```
+     Remove-Item -Recurse -Force .githooks, .github/workflows/rodo.yml, .claude/hooks/rodo-guard.cjs, scripts/rodo-lint.cjs, scripts/rodo-lint, scripts/rodo-done.cjs, scripts/rodo-approval.cjs -ErrorAction SilentlyContinue
+     ```
+   - Linux/macOS:
+     ```
+     rm -rf .githooks .github/workflows/rodo.yml .claude/hooks/rodo-guard.cjs scripts/rodo-lint.cjs scripts/rodo-lint scripts/rodo-done.cjs scripts/rodo-approval.cjs
+     ```
+   - `.claude/settings.json`: if it only contains the RODO hooks (`rodo-guard.cjs`), delete it; otherwise remove just the two `rodo-guard.cjs` entries.
+   - Delete `rodo/cycles/*/acceptance-result.json` (no longer used).
+3. **Config:** in `rodo/config.json`, rename `acceptanceCommand` to `testCommand` and delete `acceptanceItem`, `ciTestCommand`, `approvalRequiresSync`.
+4. **Cards:**
+   - `[approval: <hash>]` → `[approval]` (or remove the approval line if you do not want approval for that card).
+   - Statuses stay five: `todo`, `in_progress`, `blocked`, `cancelled`, `done`. A `blocked` card needs its reason on `- Blocker:` in `## Progress notes`; a `cancelled` card is kept (not deleted) with `- Cancelled: <reason>` in Progress notes and a line under `## لغوشده` in MAP.md. Add these lines to old cards that lack them.
+   - The RODO board needs a server that knows `blocked` and `cancelled` (RODO cycle 004); older servers report an error for those two statuses (nothing else is blocked).
+   - Delete `x-failures` lines.
+5. **CHANGELOG:** new lines use `- YYYY-MM-DD | what | why | files` (old 3-part lines can stay).
+6. Compare your `rodo/RULES.md` with the kit's `rodo/RULES.md` and merge by hand (it is never replaced). The new decision rule for "card or CHANGELOG line" is in section 4.
+7. Commit: `git add -A && git commit -m "chore: RODO kit 2.0.0 (method only)"`.
+
+## 1.0.0 → 1.1.0
+
+Superseded by 2.0.0. If you are still on 1.0.0, go straight to 2.0.0: do the steps above (step 1 also covers `scripts/hooks`: delete that folder too).

package/dist/kit/agents-block.md

@@ -0,0 +1,6 @@
+<!-- rodo:start -->
+This repo uses the RODO method (kit 2.2.0). Before any work: read `.agents/skills/rodo/SKILL.md`,
+run `rodo sync`, then `rodo resume`. At the end of every session run `rodo sync`.
+RODO projects: see `rodo/*/project.json`. Never write secrets into rodo files. Never create boards.
+Never run `git pull`, `git fetch`, `git merge` or `git rebase`. If the repo is behind, tell the user once.
+<!-- rodo:end -->

package/dist/kit/kit-manifest.json

@@ -0,0 +1,228 @@
+{
+  "version": "2.2.0",
+  "files": [
+    {
+      "src": ".agents/skills/rodo/SKILL.md",
+      "dest": ".agents/skills/rodo/SKILL.md",
+      "owner": "kit"
+    },
+    {
+      "src": ".agents/skills/rodo/coder.md",
+      "dest": ".agents/skills/rodo/coder.md",
+      "owner": "kit"
+    },
+    {
+      "src": ".agents/skills/rodo/onboarding.md",
+      "dest": ".agents/skills/rodo/onboarding.md",
+      "owner": "kit"
+    },
+    {
+      "src": ".agents/skills/rodo/planner.md",
+      "dest": ".agents/skills/rodo/planner.md",
+      "owner": "kit"
+    },
+    {
+      "src": ".agents/skills/rodo/small-change.md",
+      "dest": ".agents/skills/rodo/small-change.md",
+      "owner": "kit"
+    },
+    {
+      "src": ".agents/skills/rodo/templates/acceptance.feature",
+      "dest": ".agents/skills/rodo/templates/acceptance.feature",
+      "owner": "kit"
+    },
+    {
+      "src": ".agents/skills/rodo/templates/card.rodo.md",
+      "dest": ".agents/skills/rodo/templates/card.rodo.md",
+      "owner": "kit"
+    },
+    {
+      "src": ".agents/skills/rodo/templates/design.md",
+      "dest": ".agents/skills/rodo/templates/design.md",
+      "owner": "kit"
+    },
+    {
+      "src": ".agents/skills/rodo/templates/en/acceptance.feature",
+      "dest": ".agents/skills/rodo/templates/en/acceptance.feature",
+      "owner": "kit"
+    },
+    {
+      "src": ".agents/skills/rodo/templates/en/card.rodo.md",
+      "dest": ".agents/skills/rodo/templates/en/card.rodo.md",
+      "owner": "kit"
+    },
+    {
+      "src": ".agents/skills/rodo/templates/en/request.rodo.md",
+      "dest": ".agents/skills/rodo/templates/en/request.rodo.md",
+      "owner": "kit"
+    },
+    {
+      "src": ".agents/skills/rodo/templates/en/simple.md",
+      "dest": ".agents/skills/rodo/templates/en/simple.md",
+      "owner": "kit"
+    },
+    {
+      "src": ".agents/skills/rodo/templates/request.rodo.md",
+      "dest": ".agents/skills/rodo/templates/request.rodo.md",
+      "owner": "kit"
+    },
+    {
+      "src": ".agents/skills/rodo/templates/simple.md",
+      "dest": ".agents/skills/rodo/templates/simple.md",
+      "owner": "kit"
+    },
+    {
+      "src": ".agents/skills/rodo/templates/spec.md",
+      "dest": ".agents/skills/rodo/templates/spec.md",
+      "owner": "kit"
+    },
+    {
+      "src": ".agents/skills/rodo/testing.md",
+      "dest": ".agents/skills/rodo/testing.md",
+      "owner": "kit"
+    }
+  ],
+  "legacyHashes": {
+    "AGENTS.md": [
+      "sha256:936df4b30128738e224072e1c3aebbe8899793e9582da97966a64c13a01395e4"
+    ],
+    "CLAUDE.md": [
+      "sha256:336cc4fbf19beaada7ccf9986414fa91851a8d7a07dfb3ccbe800a69eed0ab49"
+    ],
+    ".agents/skills/rodo/SKILL.md": [
+      "sha256:a06382e130c46d0d7fc8ebfed4610a829496be1d0830ba79bf3135fa2da7b05b"
+    ],
+    ".agents/skills/rodo/coder.md": [
+      "sha256:6a5b3632e4c3b8065e8611fea5c267719bd9eddc797a7987723f78ef055e49b4"
+    ],
+    ".agents/skills/rodo/onboarding.md": [
+      "sha256:cd7cd64e646ff1762dd3b7fa27a6441f76222a3ad153d092f6120c9d9d11d0de"
+    ],
+    ".agents/skills/rodo/planner.md": [
+      "sha256:36e6082c88920c2bf03f51e2e1a93f92cd2d06406d42a1cda357a65ca85f89e3"
+    ],
+    ".agents/skills/rodo/small-change.md": [
+      "sha256:7bb7ceb277773e1ec4ed351ce0591e39dc5ba0911e0a32cce358ae16e2358c12"
+    ],
+    ".agents/skills/rodo/templates/CHANGELOG.md": [
+      "sha256:74b575d434805ef245587bf704f0cda130d2cd0fb37e2499834853e0f372a059"
+    ],
+    ".agents/skills/rodo/templates/MAP.md": [
+      "sha256:9597d20bc5871cd5699f03e18b16356d06f212de37afe52ddfd3fd3f3e7ed200"
+    ],
+    ".agents/skills/rodo/templates/MEMORY.md": [
+      "sha256:68920278b000efa6b64cc206553720e14cb8af3a4154623e5babef3a745c0e03"
+    ],
+    ".agents/skills/rodo/templates/acceptance.feature": [
+      "sha256:0156ea9eb06f09de596a52a5ad2b1852b351d9cbe7a28230a4f90a2138a4b131"
+    ],
+    ".agents/skills/rodo/templates/architecture.md": [
+      "sha256:ecf9d97143671824620c74c9d3d0e19514b242a4b6c1552440802a198d7e2972"
+    ],
+    ".agents/skills/rodo/templates/brief.md": [
+      "sha256:085f4338ac8cb8c6b9afb1fd22a5a8aed9cf5694f2a2183e47a1fac6a0db9490"
+    ],
+    ".agents/skills/rodo/templates/card.rodo.md": [
+      "sha256:6639b0b83bf25d8355f576b367e8fa10b51b0f5e7f100b5257c78cb4d3d0cc6a"
+    ],
+    ".agents/skills/rodo/templates/config.json": [
+      "sha256:0c07d153aad92e647a8211a54cf696cfad5ac1445593383738da4cf29108f669"
+    ],
+    ".agents/skills/rodo/templates/glossary.md": [
+      "sha256:0a5b2eff02a30ad5753547929789c8f3a6e39158ab65b433198e9845df42dfb6"
+    ],
+    ".agents/skills/rodo/templates/simple.md": [
+      "sha256:b81908a8886698cce148400113a2525f6c44cf34c8e412290ba1652b3b928b6f"
+    ],
+    ".agents/skills/rodo/templates/spec.md": [
+      "sha256:3f9f0bcbdf75f7a1981eb52efdbc2bab90107161af83cd342a2de514b56e1a25"
+    ],
+    ".agents/skills/rodo/testing.md": [
+      "sha256:0cb154fdf25dc5222a63973ad97d7e02a3c3093b800f0072eb1fd58423c32c41"
+    ]
+  },
+  "projectTemplate": [
+    {
+      "src": "project-template/CHANGELOG.md",
+      "dest": "CHANGELOG.md",
+      "owner": "project"
+    },
+    {
+      "src": "project-template/MAP.md",
+      "dest": "MAP.md",
+      "owner": "project"
+    },
+    {
+      "src": "project-template/MEMORY.md",
+      "dest": "MEMORY.md",
+      "owner": "project"
+    },
+    {
+      "src": "project-template/RULES.md",
+      "dest": "RULES.md",
+      "owner": "project"
+    },
+    {
+      "src": "project-template/project.json",
+      "dest": "project.json",
+      "owner": "project"
+    },
+    {
+      "src": "project-template/project/architecture.md",
+      "dest": "project/architecture.md",
+      "owner": "project"
+    },
+    {
+      "src": "project-template/project/brief.md",
+      "dest": "project/brief.md",
+      "owner": "project"
+    },
+    {
+      "src": "project-template/project/glossary.md",
+      "dest": "project/glossary.md",
+      "owner": "project"
+    }
+  ],
+  "projectTemplateEn": [
+    {
+      "src": "project-template-en/CHANGELOG.md",
+      "dest": "CHANGELOG.md",
+      "owner": "project"
+    },
+    {
+      "src": "project-template-en/MAP.md",
+      "dest": "MAP.md",
+      "owner": "project"
+    },
+    {
+      "src": "project-template-en/MEMORY.md",
+      "dest": "MEMORY.md",
+      "owner": "project"
+    },
+    {
+      "src": "project-template-en/RULES.md",
+      "dest": "RULES.md",
+      "owner": "project"
+    },
+    {
+      "src": "project-template-en/project.json",
+      "dest": "project.json",
+      "owner": "project"
+    },
+    {
+      "src": "project-template-en/project/architecture.md",
+      "dest": "project/architecture.md",
+      "owner": "project"
+    },
+    {
+      "src": "project-template-en/project/brief.md",
+      "dest": "project/brief.md",
+      "owner": "project"
+    },
+    {
+      "src": "project-template-en/project/glossary.md",
+      "dest": "project/glossary.md",
+      "owner": "project"
+    }
+  ]
+}

package/dist/kit/project-template/CHANGELOG.md

@@ -0,0 +1,6 @@
+# تغییرهای کوچک
+
+> هر تغییر کوچکی که رفتار توضیح‌داده‌شده در کارت‌ها، spec یا سناریوها را عوض نمی‌کند، یک خط (بدون کارت).
+> قالب: `- YYYY-MM-DD | چه چیزی عوض شد | چرا | فایل‌ها`
+
+- {{DATE}} | روش RODO به پروژه اضافه شد | تا AI و آدم‌ها از یک نقشه و سندهای به‌روز کار کنند | `rodo/`

package/dist/kit/project-template/MAP.md

@@ -0,0 +1,22 @@
+# نقشهٔ پروژه: {{نام پروژه}}
+
+> هر کارت و سندهایش در `rodo/<project>/cycles/NNN-slug/` می‌ماند. این فایل وقتی یک کارت تمام یا لغو می‌شود به‌روز می‌شود (نه وسط کدنویسی)، تا چند نفر هم‌زمان کار کنند و Git تداخل نگیرد. وضعیت کار جاری در کارت‌هاست.
+> کوتاه بماند (حدود ۴۰ خط). اول از همه همین را بخوانید.
+
+## این پروژه چیست
+
+{{یک یا دو جمله به زبان ساده}}
+
+## الان کجاییم
+
+{{یک یا دو جمله: آخرین چیزی که تمام شد}}
+
+## انجام‌شده
+
+## لغوشده
+
+> قالب هر خط: لینک چرخه، بعد « — »، بعد دلیل لغو.
+
+## بعدی
+
+- {{ایدهٔ بعدی}}

package/dist/kit/project-template/MEMORY.md

@@ -0,0 +1,6 @@
+# حافظهٔ تصمیم‌ها
+
+> هر تصمیم مهم یک خط: چه تصمیمی، چون چه. وقتی یک کارت تمام یا لغو می‌شود اضافه می‌شود. تصمیم‌ها پاک نمی‌شوند؛ اگر عوض شدند، خط تازه بنویسید و بگویید جای کدام را می‌گیرد.
+> قالب: `- YYYY-MM-DD — تصمیم. چون دلیل. (cycle NNN)`
+
+- {{DATE}} — کار پروژه با روش RODO پیش می‌رود. چون AI و آدم‌ها باید از یک نقشه، کارت‌ها و سندهای به‌روز کار کنند.