[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M74F5YK8SVG6QXT9DDYHTJZP":3,"skill-same:01M74F5YK8SVG6QXT9DDYHTJZP":62},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":15,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":32,"commitSha":33,"skillDirectory":34,"forksCount":13,"repositoryHtmlUrl":35,"repositoryHomepageUrl":6,"ownerHtmlUrl":36,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":37,"install":38,"resources":51},"01M74F5YK8SVG6QXT9DDYHTJZP","mem0-test-integration",null,"Verify a Mem0 integration produced by \u002Fmem0-integrate. Runs in the same workspace on the same branch (loose coupling) — installs dependencies, runs the repo's native test suite, then exercises a real end-to-end smoke flow against the user's API key. Produces a scorecard. TRIGGER when: user has just run \u002Fmem0-integrate and says \"verify\", \"test the integration\", \"run \u002Fmem0-test-integration\", or when a .mem0-integration\u002F directory exists and tests have not been run yet on the current branch. DO NOT TRIGGER when: the user wants to run general project tests (defer to the repo's native test command), or when no prior \u002Fmem0-integrate run exists in the current branch (ask them to run \u002Fmem0-integrate first). This skill ONLY catches compile and runtime bugs by design. Logical integration errors — wrong data stored, wrong time retrieved, wrong user scoping — are on the human reviewer.","lzy18001500226\u002FMoSim","lzy18001500226","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F152623587?v=4","01M71ZMEX36NK05B642131DEEK","01M71ZMEXSMX72KPVNJ572G0CM",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002FReferences\u002FAgent\u002FMemory\u002Fmem0\u002Fskills\u002Fmem0-test-integration\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002FReferences\u002FAgent\u002FMemory\u002Fmem0\u002Fskills\u002Fmem0-test-integration\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: mem0-test-integration\ndescription: >\n  Verify a Mem0 integration produced by \u002Fmem0-integrate. Runs in the same\n  workspace on the same branch (loose coupling) — installs dependencies,\n  runs the repo's native test suite, then exercises a real end-to-end\n  smoke flow against the user's API key. Produces a scorecard.\n  TRIGGER when: user has just run \u002Fmem0-integrate and says \"verify\",\n  \"test the integration\", \"run \u002Fmem0-test-integration\", or when a\n  .mem0-integration\u002F directory exists and tests have not been run yet\n  on the current branch.\n  DO NOT TRIGGER when: the user wants to run general project tests\n  (defer to the repo's native test command), or when no prior \u002Fmem0-integrate\n  run exists in the current branch (ask them to run \u002Fmem0-integrate first).\n  This skill ONLY catches compile and runtime bugs by design. Logical\n  integration errors — wrong data stored, wrong time retrieved, wrong\n  user scoping — are on the human reviewer.\nlicense: Apache-2.0\nmetadata:\n  author: mem0ai\n  version: \"0.1.0\"\n  category: ai-memory\n  tags: \"memory, integration, testing, tdd, platform, oss\"\n  coupling: loose\n  mem0_tested_versions: \"mem0ai (PyPI) >=2.0.0,\u003C3.0.0; mem0ai (npm) >=3.0.0,\u003C4.0.0\"\n---\n\n# mem0-test-integration\n\nVerifies what `\u002Fmem0-integrate` produced. Runs in the same workspace,\non the same feature branch. Loose coupling — fast, catches compile and\nruntime bugs, does not catch logical errors.\n\n## Canonical sources (use these, not ambient knowledge)\n\nAll static checks and smoke-test shapes validate against these URLs.\n`WebFetch` each before running step 3.\n\n- Scope-tagged docs index: https:\u002F\u002Fdocs.mem0.ai\u002Fllms.txt\n- OpenAPI (Platform REST): https:\u002F\u002Fdocs.mem0.ai\u002Fopenapi.json\n- Published SDK skill (canonical call patterns): https:\u002F\u002Fraw.githubusercontent.com\u002Fmem0ai\u002Fmem0\u002Fmain\u002Fskills\u002Fmem0\u002FSKILL.md\n- Vercel AI SDK skill (if the target repo uses `@ai-sdk\u002F*`): https:\u002F\u002Fraw.githubusercontent.com\u002Fmem0ai\u002Fmem0\u002Fmain\u002Fskills\u002Fmem0-vercel-ai-sdk\u002FSKILL.md\n- SDK source (cross-check version against frontmatter `mem0_tested_versions`):\n  - Repo root: https:\u002F\u002Fgithub.com\u002Fmem0ai\u002Fmem0\n  - Python: https:\u002F\u002Fgithub.com\u002Fmem0ai\u002Fmem0\u002Ftree\u002Fmain\u002Fmem0\n  - TypeScript: https:\u002F\u002Fgithub.com\u002Fmem0ai\u002Fmem0\u002Ftree\u002Fmain\u002Fmem0-ts\n\nRead the `Delegated skill:` field in `.mem0-integration\u002Fplan.md` — if it\nnames a skill URL, fetch that skill and use its example blocks as the\nreference for both static checks (step 3) and the smoke test (step 5).\n\n## Non-invasiveness contract\n\nEvery check in this skill assumes the integration is **additive and\nfeature-flagged** (see `\u002Fmem0-integrate` \"Integration principles\").\nSpecifically:\n\n- `product.json` must contain a `feature_flag` field.\n- Steps 4–6 run in two passes:\n  - **Pass A — flag unset.** All pre-existing tests must pass, smoke\u002FE2E\n    skip. The repo must behave like `main`. Any failure here is a\n    **hard fail** — do not let the self-heal loop attempt a patch.\n  - **Pass B — flag set.** New tests must pass, smoke and E2E run.\n- If Pass A fails, the scorecard marks `non_invasive: false` and sets\n  `overall: fail` with a distinct reason code the integrator's heal\n  loop refuses to touch.\n\n## Preconditions\n\nRefuse to start unless ALL of the following are true:\n\n- `.mem0-integration\u002F` directory exists in the repo root.\n- `.mem0-integration\u002Fproduct.json`, `goal.md`, and `plan.md` are readable\n  and internally consistent (JSON parses, docs non-empty).\n- Current branch name begins with `mem0-integrate\u002F` (set by the companion\n  skill). Prevents accidental runs on unrelated branches.\n- Working tree is clean. The skill never modifies source files; any dirty\n  state means the integration is mid-edit and not ready to verify.\n- The same API key the integration used is available in the environment\n  (`MEM0_API_KEY` for Platform, `OPENAI_API_KEY` for OSS — read which from\n  `product.json`). Interactive mode asks if missing; CI mode exits 2.\n\nExit with a written rationale on any precondition failure. Never attempt\nto \"fix up\" state.\n\n## Pipeline\n\n### 1. Read the contract\n\nLoad:\n\n- `product.json` → which language, which product (Platform vs OSS), which\n  mem0 version, `write_site`, `read_site`.\n- `plan.md` → the mechanical contract (write pattern, read pattern,\n  preserved behavior).\n- `goal.md` → the intent (displayed in the scorecard only; not tested).\n\n### 2. Install dependencies\n\nRoute by language from `product.json`:\n\n| Language | Command |\n|---|---|\n| Python | `pip install -e .` if editable, else `pip install -r requirements.txt`. Then `pip install mem0ai` if not already present at the pinned version. |\n| TypeScript \u002F JavaScript | `npm install` (or `pnpm install` \u002F `yarn install` if detected by lockfile). |\n\nIf install fails → exit code 2 with stderr tail. Never move to testing\nif dependencies don't resolve.\n\n### 3. Static sanity checks (fast, local, no API calls)\n\n- **Import check**: does the write-site file import the expected Mem0\n  surface? Authoritative list comes from `## Identify the User's Setup`\n  in `https:\u002F\u002Fdocs.mem0.ai\u002Fllms.txt`:\n  - Platform Python → `from mem0 import MemoryClient`\n  - Platform TS → `import MemoryClient from \"mem0ai\"`\n  - OSS Python → `from mem0 import Memory`\n  - OSS TS → `import { Memory } from \"mem0ai\u002Foss\"`\n\n  If `plan.md` names a delegated skill (e.g., Vercel AI), use *that*\n  skill's import signature instead of the list above. Mismatch → fail\n  with line number.\n- **Version check**: installed `mem0ai` version falls in the range from\n  this skill's `mem0_tested_versions`. Out of range → warn but continue.\n- **Type check** (TS tracks only): run `tsc --noEmit` or `tsup --dts`.\n  Non-zero → fail.\n- **Lint** (if the repo has a linter configured): run the repo's own\n  lint command. Lint failures from this skill's changes → fail; pre-existing\n  lint failures → surface as a warning.\n- **Eager-init check**: grep the `write_site` and `read_site` files (paths\n  from `product.json`) for `MemoryClient(` or `Memory(` at module scope —\n  i.e., not inside a function, method, or class body. `MemoryClient()`\n  validates the API key in `__init__` (network call) and OSS `Memory()`\n  can eagerly initialize embedding\u002FLLM providers — module-level\n  instantiation hits the wire on import and breaks Pass A's test\n  collection whenever the key is unset. Hit → fail with `file:line` and\n  the lazy-init guidance from `\u002Fmem0-integrate` step 8 constraint #7.\n\n### 4. Run the repo's native test suite (two passes)\n\n| Language | Test command (in priority order) |\n|---|---|\n| Python | `pytest` with the test files from step 5 of the companion skill, else `python -m unittest discover`. |\n| TypeScript \u002F JavaScript | `npm test` if defined in package.json; else auto-detect `vitest` or `jest`. |\n\n**Pass A — `feature_flag` unset.** Run the *entire* pre-existing suite\n(excluding the new `test_mem0_*` files). **Must be 100% green.** Any\nfailure here marks `non_invasive: false` in the scorecard and is\na **hard fail** — the integrator's self-heal loop refuses to touch it.\n\n**Pass B — `feature_flag` set** (value from `product.json`). Run the\nfull suite including the new tests. All must pass.\n\nIsolate integration-introduced failures using `git diff main..HEAD\n--name-only`. A test file that exists on `main` and fails only under\nthe integration branch (flag set *or* unset) counts against the\nscorecard regardless of pass. A test file that already failed on `main`\nis surfaced as `pre_existing_unrelated` and does not count — but is\nstill reported so the user can clean it up.\n\nCapture output to `.mem0-integration\u002Ftest-stdout-flag-off.log` and\n`.mem0-integration\u002Ftest-stdout-flag-on.log`. Scorecard reports pass\u002Ffail\nper pass.\n\n### 5. Smoke test (real API call, shortest round-trip)\n\nScripted end-to-end flow tailored to `product.json`. The call shapes\nbelow are the minimal ones; if `plan.md` names a delegated skill, use\n*that skill's* minimal example verbatim instead — it is the canonical\nshape for the detected stack.\n\n**Platform (Python):**\n\n    from mem0 import MemoryClient\n    c = MemoryClient()                               # uses MEM0_API_KEY\n    uid = f\"mem0-test-integration-{os.urandom(4).hex()}\"\n    c.add([{\"role\": \"user\", \"content\": \"I prefer aisle seats\"}], user_id=uid)\n    hits = c.search(\"seat preference\", user_id=uid)\n    assert any(\"aisle\" in h.get(\"memory\", \"\") for h in hits), hits\n    c.delete_all(user_id=uid)                        # clean up\n\n**Platform (TS):** same shape with `MemoryClient` from `\"mem0ai\"`.\n\n**OSS (Python \u002F TS):** uses `Memory()` \u002F `new Memory()` with default config\n(OpenAI LLM via `OPENAI_API_KEY`, local Qdrant). If the repo ships a\n`docker-compose.yml` with a Qdrant service, the skill starts it first and\ntears it down after. If no backing store is reachable → fail with a\nclear message naming the fix.\n\nThe smoke test always uses a **disposable random user_id** prefixed with\n`mem0-test-integration-` so a failed cleanup doesn't pollute the user's\nreal data. A background tidy step deletes any prefix-matching entries\nolder than 24 hours on the next run.\n\nCapture output to `.mem0-integration\u002Fsmoke-stdout.log`.\n\n### 6. E2E integration test (run the app, exercise the flow)\n\nUnit tests + smoke prove the SDK works in isolation. This step is the\nreal signal: **does memory actually appear in the app's user-visible\noutput when the integration runs end-to-end?**\n\nRequires `plan.md` to contain an `E2E recipe:` section (authored by\n`\u002Fmem0-integrate` step 5). If absent → status `skipped` (not `fail`),\nnote in scorecard that the repo has no runnable entry point.\n\nRecipe fields the skill reads:\n\n- `start` — shell command to launch the app using `$PORT` for any network\n  port. Run in background with stdout\u002Fstderr teed to\n  `.mem0-integration\u002Fe2e-app.log`.\n- `ready_probe` — how to detect readiness. `url=... status=...` polls an\n  HTTP endpoint; `log=\"...\"` waits for a substring in `e2e-app.log`;\n  `sleep=N` waits N seconds (last resort). 60-second hard timeout.\n- `compose_services` — optional. If set, bring them up via\n  `docker compose up -d \u003Cservices>` before `start`, tear them down with\n  `docker compose down` at the end.\n- `write_call` — triggers the Mem0 write path exactly once. Output is\n  captured and surfaced on failure. 60-second hard timeout.\n- `write_async_wait_ms` — pause after `write_call` to let async memory\n  flushes land. Default 0.\n- `read_call` — triggers the Mem0 read path. Typically a fresh session\n  or new request that should surface the stored memory.\n- `read_assert` — substring, `regex=...`, or `jsonpath=\u003Cexpr>=\u003Cvalue>`\n  that must appear in `read_call`'s stdout. This is the E2E pass gate.\n\nExecution order:\n\n1. Allocate an ephemeral TCP port; export as `PORT`.\n2. Set `MEM0_USER_ID` to a disposable `mem0-test-integration-\u003Crand>` value\n   and export it, so the app can use the same scoping the smoke test does\n   if the recipe wants cleanup.\n3. Bring up `compose_services` if named.\n4. Run `start` in the background.\n5. Poll `ready_probe` until success or 60s timeout. Timeout → fail.\n6. Run `write_call`. Non-zero exit → fail (but continue to cleanup).\n7. Sleep `write_async_wait_ms`.\n8. Run `read_call`.\n9. Evaluate `read_assert` against `read_call`'s stdout. Miss → fail.\n10. Cleanup (always, even on failure): SIGTERM the app, SIGKILL after\n    5s, `docker compose down` if services were started, `delete_all`\n    memories matching `mem0-test-integration-*` on Platform scenarios.\n\nOn any failure, the scorecard includes:\n\n- Last 40 lines of `e2e-app.log`\n- Full `write_call` output\n- Full `read_call` output\n- The expected vs actual for `read_assert`\n\n### 7. Scorecard\n\nWrite `.mem0-integration\u002Fscorecard.md` and `.mem0-integration\u002Fscorecard.json`:\n\n    {\n      \"timestamp\": \"2026-04-20T14:03:11Z\",\n      \"branch\": \"mem0-integrate\u002Fremember-user-preferences\",\n      \"product\": \"platform\",\n      \"language\": \"python\",\n      \"mem0_version\": \"2.0.0\",\n      \"non_invasive\": true,\n      \"feature_flag\": \"MEM0_ENABLED\",\n      \"results\": {\n        \"install\":      {\"status\": \"pass\", \"duration_ms\": 12043},\n        \"static_checks\":{\"status\": \"pass\", \"duration_ms\": 812},\n        \"unit_tests_flag_off\": {\"status\": \"pass\", \"duration_ms\": 3920, \"count\": 47,\n                                \"reason\": \"all pre-existing tests green with flag unset\"},\n        \"unit_tests_flag_on\":  {\"status\": \"pass\", \"duration_ms\": 4321, \"count\": 49},\n        \"smoke_test\":   {\"status\": \"pass\", \"duration_ms\": 2890, \"memory_id\": \"mem_...\"},\n        \"e2e_test\":     {\"status\": \"pass\", \"duration_ms\": 14200,\n                         \"ready_probe_ms\": 3100, \"write_exit\": 0,\n                         \"read_assert_matched\": true}\n      },\n      \"friction\": {\n        \"dependency_install_retries\": 0,\n        \"pre_existing_test_failures\": 0,\n        \"warnings\": [\"mem0ai 2.0.0 pinned; consider 2.0.1 for fix X\"]\n      },\n      \"overall\": \"pass\"\n    }\n\nThe markdown version is human-readable and includes:\n\n- Goal doc + plan doc reprinted at top (so reviewers don't have to hunt).\n- Each check with pass\u002Ffail + log excerpt.\n- Friction summary.\n- Verbatim warnings from mem0 SDK (if any — e.g., deprecated field usage).\n- **Explicit \"NOT checked\" section** listing what loose coupling misses:\n  \"Whether the stored data is what the user wants stored. Whether search\n  runs at the right moment. Whether user_id matches the actual session\n  scope. Human review required.\"\n\n### 8. Report + exit\n\n- Print the scorecard path + overall pass\u002Ffail to stdout.\n- **Do not commit the scorecard files.** They live in `.mem0-integration\u002F`,\n  which is gitignored. The user can inspect and optionally pin.\n- On fail: print the first failing step's log tail (last 40 lines) and\n  stop. Do not attempt to fix anything.\n\n## Artifacts (all under `.mem0-integration\u002F`)\n\n| File | Purpose | Retention |\n|---|---|---|\n| `scorecard.md` | Human-readable verdict. | Overwritten per run. |\n| `scorecard.json` | Machine-readable verdict. Consumed by the CI scorecard workflow later. | Overwritten per run. |\n| `test-stdout-flag-off.log` | Step 4 Pass A (pre-existing suite, flag unset). | Overwritten per run. |\n| `test-stdout-flag-on.log` | Step 4 Pass B (full suite, flag set). | Overwritten per run. |\n| `smoke-stdout.log` | Full output from step 5. | Overwritten per run. |\n| `e2e-app.log` | Background app stdout\u002Fstderr from step 6. | Overwritten per run. |\n| `e2e-calls.log` | write_call + read_call invocations and outputs. | Overwritten per run. |\n\n## Modes\n\n| Mode | Trigger | Behavior |\n|---|---|---|\n| Interactive (default) | TTY present, `MEM0_TEST_CI` unset | Asks for missing keys, prints friendly summaries. |\n| CI | `MEM0_TEST_CI=1` | Keys must be in env, no prompts, non-zero exit on any fail. JSON scorecard goes to stdout's tail for workflow parsing. |\n\n## Invocation\n\n    \u002Fmem0-test-integration                       # interactive, all steps\n    \u002Fmem0-test-integration --ci                  # non-interactive\n    \u002Fmem0-test-integration --skip-smoke          # no API calls, no E2E\n    \u002Fmem0-test-integration --skip-e2e            # unit + smoke only (faster CI)\n    \u002Fmem0-test-integration --only-smoke          # just smoke\n    \u002Fmem0-test-integration --only-e2e            # just E2E (assumes deps installed)\n\nComposition: `--skip-*` can stack (`--skip-smoke --skip-e2e` = static +\nunit only, zero API cost). `--only-*` is mutually exclusive with all\nother flags.\n\n## Exit codes\n\n| Code | Meaning |\n|---|---|\n| 0 | All checks passed. |\n| 1 | Precondition failed (no `.mem0-integration\u002F`, wrong branch, dirty tree). |\n| 2 | Missing env key (CI mode) or dependency install failure. |\n| 3 | Static sanity check failed (wrong import, type error). |\n| 4 | Unit tests failed (Pass B — integration itself broken). |\n| 5 | Smoke test failed. |\n| 6 | E2E test failed (ready_probe timeout, write\u002Fread call failed, or read_assert miss). |\n| 7 | Non-invasiveness violation: Pass A failed (pre-existing tests broke). Integrator's heal loop refuses to touch this. |\n| 8 | Internal error (skill bug — report it). |\n\n## Explicitly out of scope\n\n- **Modifying source files.** The skill is read-only against the repo.\n  If verification exposes a bug, re-run `\u002Fmem0-integrate` on the same\n  goal + plan; do not hand-patch.\n- **Fixing broken tests.** Failing unit tests are a signal that the\n  integration is wrong, not that the tests are wrong. The skill does\n  not \"try a different test.\"\n- **Deep logical correctness.** The E2E step proves \"something the user\n  said earlier comes back later,\" which is a useful but shallow signal.\n  It does NOT prove the integration picks the *right* facts to store,\n  scopes `user_id` correctly across real users, or handles conflict\n  resolution well. That's human review territory.\n- **Self-healing.** This skill never modifies source files. The paired\n  `\u002Fmem0-integrate` skill in its default `--heal` mode consumes the\n  scorecard produced here and drives its own remediation loop. Exit\n  code 7 (non-invasiveness violation) is the explicit signal the heal\n  loop must stop and surface to the user.\n- **Cross-branch comparisons.** No `main` baseline diffing. The\n  scorecard reflects this branch only.\n- **Running against production data.** Every smoke test uses a disposable\n  random user_id and cleans up after. Never touches any other user's data.\n","# mem0-test-integration\n\nVerifies what `\u002Fmem0-integrate` produced. Runs in the same workspace,\non the same feature branch. Loose coupling — fast, catches compile and\nruntime bugs, does not catch logical errors.\n\n## Canonical sources (use these, not ambient knowledge)\n\nAll static checks and smoke-test shapes validate against these URLs.\n`WebFetch` each before running step 3.\n\n- Scope-tagged docs index: https:\u002F\u002Fdocs.mem0.ai\u002Fllms.txt\n- OpenAPI (Platform REST): https:\u002F\u002Fdocs.mem0.ai\u002Fopenapi.json\n- Published SDK skill (canonical call patterns): https:\u002F\u002Fraw.githubusercontent.com\u002Fmem0ai\u002Fmem0\u002Fmain\u002Fskills\u002Fmem0\u002FSKILL.md\n- Vercel AI SDK skill (if the target repo uses `@ai-sdk\u002F*`): https:\u002F\u002Fraw.githubusercontent.com\u002Fmem0ai\u002Fmem0\u002Fmain\u002Fskills\u002Fmem0-vercel-ai-sdk\u002FSKILL.md\n- SDK source (cross-check version against frontmatter `mem0_tested_versions`):\n  - Repo root: https:\u002F\u002Fgithub.com\u002Fmem0ai\u002Fmem0\n  - Python: https:\u002F\u002Fgithub.com\u002Fmem0ai\u002Fmem0\u002Ftree\u002Fmain\u002Fmem0\n  - TypeScript: https:\u002F\u002Fgithub.com\u002Fmem0ai\u002Fmem0\u002Ftree\u002Fmain\u002Fmem0-ts\n\nRead the `Delegated skill:` field in `.mem0-integration\u002Fplan.md` — if it\nnames a skill URL, fetch that skill and use its example blocks as the\nreference for both static checks (step 3) and the smoke test (step 5).\n\n## Non-invasiveness contract\n\nEvery check in this skill assumes the integration is **additive and\nfeature-flagged** (see `\u002Fmem0-integrate` \"Integration principles\").\nSpecifically:\n\n- `product.json` must contain a `feature_flag` field.\n- Steps 4–6 run in two passes:\n  - **Pass A — flag unset.** All pre-existing tests must pass, smoke\u002FE2E\n    skip. The repo must behave like `main`. Any failure here is a\n    **hard fail** — do not let the self-heal loop attempt a patch.\n  - **Pass B — flag set.** New tests must pass, smoke and E2E run.\n- If Pass A fails, the scorecard marks `non_invasive: false` and sets\n  `overall: fail` with a distinct reason code the integrator's heal\n  loop refuses to touch.\n\n## Preconditions\n\nRefuse to start unless ALL of the following are true:\n\n- `.mem0-integration\u002F` directory exists in the repo root.\n- `.mem0-integration\u002Fproduct.json`, `goal.md`, and `plan.md` are readable\n  and internally consistent (JSON parses, docs non-empty).\n- Current branch name begins with `mem0-integrate\u002F` (set by the companion\n  skill). Prevents accidental runs on unrelated branches.\n- Working tree is clean. The skill never modifies source files; any dirty\n  state means the integration is mid-edit and not ready to verify.\n- The same API key the integration used is available in the environment\n  (`MEM0_API_KEY` for Platform, `OPENAI_API_KEY` for OSS — read which from\n  `product.json`). Interactive mode asks if missing; CI mode exits 2.\n\nExit with a written rationale on any precondition failure. Never attempt\nto \"fix up\" state.\n\n## Pipeline\n\n### 1. Read the contract\n\nLoad:\n\n- `product.json` → which language, which product (Platform vs OSS), which\n  mem0 version, `write_site`, `read_site`.\n- `plan.md` → the mechanical contract (write pattern, read pattern,\n  preserved behavior).\n- `goal.md` → the intent (displayed in the scorecard only; not tested).\n\n### 2. Install dependencies\n\nRoute by language from `product.json`:\n\n| Language | Command |\n|---|---|\n| Python | `pip install -e .` if editable, else `pip install -r requirements.txt`. Then `pip install mem0ai` if not already present at the pinned version. |\n| TypeScript \u002F JavaScript | `npm install` (or `pnpm install` \u002F `yarn install` if detected by lockfile). |\n\nIf install fails → exit code 2 with stderr tail. Never move to testing\nif dependencies don't resolve.\n\n### 3. Static sanity checks (fast, local, no API calls)\n\n- **Import check**: does the write-site file import the expected Mem0\n  surface? Authoritative list comes from `## Identify the User's Setup`\n  in `https:\u002F\u002Fdocs.mem0.ai\u002Fllms.txt`:\n  - Platform Python → `from mem0 import MemoryClient`\n  - Platform TS → `import MemoryClient from \"mem0ai\"`\n  - OSS Python → `from mem0 import Memory`\n  - OSS TS → `import { Memory } from \"mem0ai\u002Foss\"`\n\n  If `plan.md` names a delegated skill (e.g., Vercel AI), use *that*\n  skill's import signature instead of the list above. Mismatch → fail\n  with line number.\n- **Version check**: installed `mem0ai` version falls in the range from\n  this skill's `mem0_tested_versions`. Out of range → warn but continue.\n- **Type check** (TS tracks only): run `tsc --noEmit` or `tsup --dts`.\n  Non-zero → fail.\n- **Lint** (if the repo has a linter configured): run the repo's own\n  lint command. Lint failures from this skill's changes → fail; pre-existing\n  lint failures → surface as a warning.\n- **Eager-init check**: grep the `write_site` and `read_site` files (paths\n  from `product.json`) for `MemoryClient(` or `Memory(` at module scope —\n  i.e., not inside a function, method, or class body. `MemoryClient()`\n  validates the API key in `__init__` (network call) and OSS `Memory()`\n  can eagerly initialize embedding\u002FLLM providers — module-level\n  instantiation hits the wire on import and breaks Pass A's test\n  collection whenever the key is unset. Hit → fail with `file:line` and\n  the lazy-init guidance from `\u002Fmem0-integrate` step 8 constraint #7.\n\n### 4. Run the repo's native test suite (two passes)\n\n| Language | Test command (in priority order) |\n|---|---|\n| Python | `pytest` with the test files from step 5 of the companion skill, else `python -m unittest discover`. |\n| TypeScript \u002F JavaScript | `npm test` if defined in package.json; else auto-detect `vitest` or `jest`. |\n\n**Pass A — `feature_flag` unset.** Run the *entire* pre-existing suite\n(excluding the new `test_mem0_*` files). **Must be 100% green.** Any\nfailure here marks `non_invasive: false` in the scorecard and is\na **hard fail** — the integrator's self-heal loop refuses to touch it.\n\n**Pass B — `feature_flag` set** (value from `product.json`). Run the\nfull suite including the new tests. All must pass.\n\nIsolate integration-introduced failures using `git diff main..HEAD\n--name-only`. A test file that exists on `main` and fails only under\nthe integration branch (flag set *or* unset) counts against the\nscorecard regardless of pass. A test file that already failed on `main`\nis surfaced as `pre_existing_unrelated` and does not count — but is\nstill reported so the user can clean it up.\n\nCapture output to `.mem0-integration\u002Ftest-stdout-flag-off.log` and\n`.mem0-integration\u002Ftest-stdout-flag-on.log`. Scorecard reports pass\u002Ffail\nper pass.\n\n### 5. Smoke test (real API call, shortest round-trip)\n\nScripted end-to-end flow tailored to `product.json`. The call shapes\nbelow are the minimal ones; if `plan.md` names a delegated skill, use\n*that skill's* minimal example verbatim instead — it is the canonical\nshape for the detected stack.\n\n**Platform (Python):**\n\n    from mem0 import MemoryClient\n    c = MemoryClient()                               # uses MEM0_API_KEY\n    uid = f\"mem0-test-integration-{os.urandom(4).hex()}\"\n    c.add([{\"role\": \"user\", \"content\": \"I prefer aisle seats\"}], user_id=uid)\n    hits = c.search(\"seat preference\", user_id=uid)\n    assert any(\"aisle\" in h.get(\"memory\", \"\") for h in hits), hits\n    c.delete_all(user_id=uid)                        # clean up\n\n**Platform (TS):** same shape with `MemoryClient` from `\"mem0ai\"`.\n\n**OSS (Python \u002F TS):** uses `Memory()` \u002F `new Memory()` with default config\n(OpenAI LLM via `OPENAI_API_KEY`, local Qdrant). If the repo ships a\n`docker-compose.yml` with a Qdrant service, the skill starts it first and\ntears it down after. If no backing store is reachable → fail with a\nclear message naming the fix.\n\nThe smoke test always uses a **disposable random user_id** prefixed with\n`mem0-test-integration-` so a failed cleanup doesn't pollute the user's\nreal data. A background tidy step deletes any prefix-matching entries\nolder than 24 hours on the next run.\n\nCapture output to `.mem0-integration\u002Fsmoke-stdout.log`.\n\n### 6. E2E integration test (run the app, exercise the flow)\n\nUnit tests + smoke prove the SDK works in isolation. This step is the\nreal signal: **does memory actually appear in the app's user-visible\noutput when the integration runs end-to-end?**\n\nRequires `plan.md` to contain an `E2E recipe:` section (authored by\n`\u002Fmem0-integrate` step 5). If absent → status `skipped` (not `fail`),\nnote in scorecard that the repo has no runnable entry point.\n\nRecipe fields the skill reads:\n\n- `start` — shell command to launch the app using `$PORT` for any network\n  port. Run in background with stdout\u002Fstderr teed to\n  `.mem0-integration\u002Fe2e-app.log`.\n- `ready_probe` — how to detect readiness. `url=... status=...` polls an\n  HTTP endpoint; `log=\"...\"` waits for a substring in `e2e-app.log`;\n  `sleep=N` waits N seconds (last resort). 60-second hard timeout.\n- `compose_services` — optional. If set, bring them up via\n  `docker compose up -d \u003Cservices>` before `start`, tear them down with\n  `docker compose down` at the end.\n- `write_call` — triggers the Mem0 write path exactly once. Output is\n  captured and surfaced on failure. 60-second hard timeout.\n- `write_async_wait_ms` — pause after `write_call` to let async memory\n  flushes land. Default 0.\n- `read_call` — triggers the Mem0 read path. Typically a fresh session\n  or new request that should surface the stored memory.\n- `read_assert` — substring, `regex=...`, or `jsonpath=\u003Cexpr>=\u003Cvalue>`\n  that must appear in `read_call`'s stdout. This is the E2E pass gate.\n\nExecution order:\n\n1. Allocate an ephemeral TCP port; export as `PORT`.\n2. Set `MEM0_USER_ID` to a disposable `mem0-test-integration-\u003Crand>` value\n   and export it, so the app can use the same scoping the smoke test does\n   if the recipe wants cleanup.\n3. Bring up `compose_services` if named.\n4. Run `start` in the background.\n5. Poll `ready_probe` until success or 60s timeout. Timeout → fail.\n6. Run `write_call`. Non-zero exit → fail (but continue to cleanup).\n7. Sleep `write_async_wait_ms`.\n8. Run `read_call`.\n9. Evaluate `read_assert` against `read_call`'s stdout. Miss → fail.\n10. Cleanup (always, even on failure): SIGTERM the app, SIGKILL after\n    5s, `docker compose down` if services were started, `delete_all`\n    memories matching `mem0-test-integration-*` on Platform scenarios.\n\nOn any failure, the scorecard includes:\n\n- Last 40 lines of `e2e-app.log`\n- Full `write_call` output\n- Full `read_call` output\n- The expected vs actual for `read_assert`\n\n### 7. Scorecard\n\nWrite `.mem0-integration\u002Fscorecard.md` and `.mem0-integration\u002Fscorecard.json`:\n\n    {\n      \"timestamp\": \"2026-04-20T14:03:11Z\",\n      \"branch\": \"mem0-integrate\u002Fremember-user-preferences\",\n      \"product\": \"platform\",\n      \"language\": \"python\",\n      \"mem0_version\": \"2.0.0\",\n      \"non_invasive\": true,\n      \"feature_flag\": \"MEM0_ENABLED\",\n      \"results\": {\n        \"install\":      {\"status\": \"pass\", \"duration_ms\": 12043},\n        \"static_checks\":{\"status\": \"pass\", \"duration_ms\": 812},\n        \"unit_tests_flag_off\": {\"status\": \"pass\", \"duration_ms\": 3920, \"count\": 47,\n                                \"reason\": \"all pre-existing tests green with flag unset\"},\n        \"unit_tests_flag_on\":  {\"status\": \"pass\", \"duration_ms\": 4321, \"count\": 49},\n        \"smoke_test\":   {\"status\": \"pass\", \"duration_ms\": 2890, \"memory_id\": \"mem_...\"},\n        \"e2e_test\":     {\"status\": \"pass\", \"duration_ms\": 14200,\n                         \"ready_probe_ms\": 3100, \"write_exit\": 0,\n                         \"read_assert_matched\": true}\n      },\n      \"friction\": {\n        \"dependency_install_retries\": 0,\n        \"pre_existing_test_failures\": 0,\n        \"warnings\": [\"mem0ai 2.0.0 pinned; consider 2.0.1 for fix X\"]\n      },\n      \"overall\": \"pass\"\n    }\n\nThe markdown version is human-readable and includes:\n\n- Goal doc + plan doc reprinted at top (so reviewers don't have to hunt).\n- Each check with pass\u002Ffail + log excerpt.\n- Friction summary.\n- Verbatim warnings from mem0 SDK (if any — e.g., deprecated field usage).\n- **Explicit \"NOT checked\" section** listing what loose coupling misses:\n  \"Whether the stored data is what the user wants stored. Whether search\n  runs at the right moment. Whether user_id matches the actual session\n  scope. Human review required.\"\n\n### 8. Report + exit\n\n- Print the scorecard path + overall pass\u002Ffail to stdout.\n- **Do not commit the scorecard files.** They live in `.mem0-integration\u002F`,\n  which is gitignored. The user can inspect and optionally pin.\n- On fail: print the first failing step's log tail (last 40 lines) and\n  stop. Do not attempt to fix anything.\n\n## Artifacts (all under `.mem0-integration\u002F`)\n\n| File | Purpose | Retention |\n|---|---|---|\n| `scorecard.md` | Human-readable verdict. | Overwritten per run. |\n| `scorecard.json` | Machine-readable verdict. Consumed by the CI scorecard workflow later. | Overwritten per run. |\n| `test-stdout-flag-off.log` | Step 4 Pass A (pre-existing suite, flag unset). | Overwritten per run. |\n| `test-stdout-flag-on.log` | Step 4 Pass B (full suite, flag set). | Overwritten per run. |\n| `smoke-stdout.log` | Full output from step 5. | Overwritten per run. |\n| `e2e-app.log` | Background app stdout\u002Fstderr from step 6. | Overwritten per run. |\n| `e2e-calls.log` | write_call + read_call invocations and outputs. | Overwritten per run. |\n\n## Modes\n\n| Mode | Trigger | Behavior |\n|---|---|---|\n| Interactive (default) | TTY present, `MEM0_TEST_CI` unset | Asks for missing keys, prints friendly summaries. |\n| CI | `MEM0_TEST_CI=1` | Keys must be in env, no prompts, non-zero exit on any fail. JSON scorecard goes to stdout's tail for workflow parsing. |\n\n## Invocation\n\n    \u002Fmem0-test-integration                       # interactive, all steps\n    \u002Fmem0-test-integration --ci                  # non-interactive\n    \u002Fmem0-test-integration --skip-smoke          # no API calls, no E2E\n    \u002Fmem0-test-integration --skip-e2e            # unit + smoke only (faster CI)\n    \u002Fmem0-test-integration --only-smoke          # just smoke\n    \u002Fmem0-test-integration --only-e2e            # just E2E (assumes deps installed)\n\nComposition: `--skip-*` can stack (`--skip-smoke --skip-e2e` = static +\nunit only, zero API cost). `--only-*` is mutually exclusive with all\nother flags.\n\n## Exit codes\n\n| Code | Meaning |\n|---|---|\n| 0 | All checks passed. |\n| 1 | Precondition failed (no `.mem0-integration\u002F`, wrong branch, dirty tree). |\n| 2 | Missing env key (CI mode) or dependency install failure. |\n| 3 | Static sanity check failed (wrong import, type error). |\n| 4 | Unit tests failed (Pass B — integration itself broken). |\n| 5 | Smoke test failed. |\n| 6 | E2E test failed (ready_probe timeout, write\u002Fread call failed, or read_assert miss). |\n| 7 | Non-invasiveness violation: Pass A failed (pre-existing tests broke). Integrator's heal loop refuses to touch this. |\n| 8 | Internal error (skill bug — report it). |\n\n## Explicitly out of scope\n\n- **Modifying source files.** The skill is read-only against the repo.\n  If verification exposes a bug, re-run `\u002Fmem0-integrate` on the same\n  goal + plan; do not hand-patch.\n- **Fixing broken tests.** Failing unit tests are a signal that the\n  integration is wrong, not that the tests are wrong. The skill does\n  not \"try a different test.\"\n- **Deep logical correctness.** The E2E step proves \"something the user\n  said earlier comes back later,\" which is a useful but shallow signal.\n  It does NOT prove the integration picks the *right* facts to store,\n  scopes `user_id` correctly across real users, or handles conflict\n  resolution well. That's human review territory.\n- **Self-healing.** This skill never modifies source files. The paired\n  `\u002Fmem0-integrate` skill in its default `--heal` mode consumes the\n  scorecard produced here and drives its own remediation loop. Exit\n  code 7 (non-invasiveness violation) is the explicit signal the heal\n  loop must stop and surface to the user.\n- **Cross-branch comparisons.** No `main` baseline diffing. The\n  scorecard reflects this branch only.\n- **Running against production data.** Every smoke test uses a disposable\n  random user_id and cleans up after. Never touches any other user's data.\n",{"name":5,"license":14,"metadata":24,"description":31},{"tags":25,"author":26,"version":27,"category":28,"coupling":29,"mem0_tested_versions":30},"memory, integration, testing, tdd, platform, oss","mem0ai","0.1.0","ai-memory","loose","mem0ai (PyPI) >=2.0.0,\u003C3.0.0; mem0ai (npm) >=3.0.0,\u003C4.0.0","Verify a Mem0 integration produced by \u002Fmem0-integrate. Runs in the same workspace on the same branch (loose coupling) — installs dependencies, runs the repo's native test suite, then exercises a real end-to-end smoke flow against the user's API key. Produces a scorecard. TRIGGER when: user has just run \u002Fmem0-integrate and says \"verify\", \"test the integration\", \"run \u002Fmem0-test-integration\", or when a .mem0-integration\u002F directory exists and tests have not been run yet on the current branch. DO NOT TRIGGER when: the user wants to run general project tests (defer to the repo's native test command), or when no prior \u002Fmem0-integrate run exists in the current branch (ask them to run \u002Fmem0-integrate first). This skill ONLY catches compile and runtime bugs by design. Logical integration errors — wrong data stored, wrong time retrieved, wrong user scoping — are on the human reviewer.\n","main","e0ec177d00f087cb04e90206b2f7d3aa2f5a1350","References\u002FAgent\u002FMemory\u002Fmem0\u002Fskills\u002Fmem0-test-integration","https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim","https:\u002F\u002Fgithub.com\u002Flzy18001500226","2026-08-31 23:29:00",{"source":8,"skillName":5,"commands":39,"prompt":49,"zipAvailable":50},[40,43,46],{"tool":41,"command":42},"npx","npx skills add lzy18001500226\u002FMoSim --skill mem0-test-integration",{"tool":44,"command":45},"bunx","bunx skills add lzy18001500226\u002FMoSim --skill mem0-test-integration",{"tool":47,"command":48},"pnpm","pnpm dlx skills add lzy18001500226\u002FMoSim --skill mem0-test-integration","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002FReferences\u002FAgent\u002FMemory\u002Fmem0\u002Fskills\u002Fmem0-test-integration\u002FSKILL.md",true,[52,56],{"path":53,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":55,"sourceHtmlUrl":15},"LICENSE","许可证",10705,{"path":57,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":60,"sourceHtmlUrl":61},"README.md","REFERENCE","参考文档",3349,"https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002FReferences\u002FAgent\u002FMemory\u002Fmem0\u002Fskills\u002Fmem0-test-integration\u002FREADME.md",[63,72,80,88,96,104],{"skillNo":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":67,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":71},"01M78RK6JGRPTD3TDSMTQWDQ9P","skill-installer","Install Codex skills into $CODEX_HOME\u002Fskills from a curated list or a GitHub repo path. Use when a user asks to list installable skills, install a curated skill, or install a skill from another repo (including private repos).","https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002Fsrc\u002FAgent\u002Fcodex-main\u002Fcodex-rs\u002Fskills\u002Fsrc\u002Fassets\u002Fsamples\u002Fskill-installer\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002Fsrc\u002FAgent\u002Fcodex-main\u002Fcodex-rs\u002Fskills\u002Fsrc\u002Fassets\u002Fsamples\u002Fskill-installer\u002FSKILL.md",[],[],"2026-11-15 14:19:17",{"skillNo":73,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":76,"sourceHtmlUrl":77,"categories":78,"keywords":79,"publishedAt":71},"01M78RK29KVRDQXSMVZX1GQTJN","skill-creator","Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Codex's capabilities with specialized knowledge, workflows, or tool integrations.","https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002Fsrc\u002FAgent\u002Fcodex-main\u002Fcodex-rs\u002Fskills\u002Fsrc\u002Fassets\u002Fsamples\u002Fskill-creator\u002Flicense.txt","https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002Fsrc\u002FAgent\u002Fcodex-main\u002Fcodex-rs\u002Fskills\u002Fsrc\u002Fassets\u002Fsamples\u002Fskill-creator\u002FSKILL.md",[],[],{"skillNo":81,"name":82,"nameZh":6,"description":83,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":84,"sourceHtmlUrl":85,"categories":86,"keywords":87,"publishedAt":71},"01M78RJPF3210Y5DA9VJ6JMMG6","openai-docs","Use for Codex models\u002Fpricing, scheduled tasks, skills, settings, setup, troubleshooting, customization, automations, and self-knowledge—including 'you,' 'your,' 'this app,' or 'this coding agent' when they refer to Codex—and for OpenAI APIs\u002Fproducts and ChatGPT Work. Also use for model choice\u002Fmigration, prompting, SDKs, Responses, Realtime, agents, evals, and Chat\u002FWork\u002FCodex comparisons. Do not use for generic app\u002Fsoftware tasks that merely mention Codex.","https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002Fsrc\u002FAgent\u002Fcodex-main\u002Fcodex-rs\u002Fskills\u002Fsrc\u002Fassets\u002Fsamples\u002Fopenai-docs\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002Fsrc\u002FAgent\u002Fcodex-main\u002Fcodex-rs\u002Fskills\u002Fsrc\u002Fassets\u002Fsamples\u002Fopenai-docs\u002FSKILL.md",[],[],{"skillNo":89,"name":90,"nameZh":6,"description":91,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":92,"sourceHtmlUrl":93,"categories":94,"keywords":95,"publishedAt":71},"01M78RJH7X67C5AY8ZVS7RQNWJ","imagegen","Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output should be a bitmap asset rather than repo-native code or vector. Do not use when the task is better handled by editing existing SVG\u002Fvector\u002Fcode-native assets, extending an established icon or logo system, or building the visual directly in HTML\u002FCSS\u002Fcanvas.","https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002Fsrc\u002FAgent\u002Fcodex-main\u002Fcodex-rs\u002Fskills\u002Fsrc\u002Fassets\u002Fsamples\u002Fimagegen\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002Fsrc\u002FAgent\u002Fcodex-main\u002Fcodex-rs\u002Fskills\u002Fsrc\u002Fassets\u002Fsamples\u002Fimagegen\u002FSKILL.md",[],[],{"skillNo":97,"name":98,"nameZh":6,"description":99,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":100,"sourceHtmlUrl":101,"categories":102,"keywords":103,"publishedAt":71},"01M78RBBAYMJ6T3ZMAYHCFM6G7","skill-creator-cc","Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, update or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.","https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002FReferences\u002FAgent\u002FWorkflow\u002Fokwinds\u002Fmiscellany\u002Fagent\u002Fskills\u002Fskill-creator-cc\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002FReferences\u002FAgent\u002FWorkflow\u002Fokwinds\u002Fmiscellany\u002Fagent\u002Fskills\u002Fskill-creator-cc\u002FSKILL.md",[],[],{"skillNo":105,"name":106,"nameZh":6,"description":107,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":108,"licenseFileUrl":109,"sourceHtmlUrl":110,"categories":111,"keywords":112,"publishedAt":113},"01M76VV9V2R9FZTVGMJZV7Y996","prd-writing-guide","Write complete, unambiguous PRDs that development teams can implement without guesswork. Includes requirement discovery framework, structured documentation methodology, completeness checklists, and common pitfall avoidance. Use when: writing new PRDs, reviewing PRD drafts, validating requirement completeness, preparing for engineering handoff. Triggers: 'write PRD', '写PRD', '产品需求文档', '需求文档', '需求规格', '需求评审', '完善需求', 'create requirements doc', 'product requirements', 'feature spec', 'requirements document'. Anti-triggers: 'technical design doc', 'architecture design', 'implementation plan', 'API design', '架构设计', '技术方案', '实现方案', '接口设计'.","MIT","https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002FReferences\u002FAgent\u002FWorkflow\u002Fokwinds\u002Fmiscellany\u002Fagent\u002Fskills\u002Fprd-writing-guide\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Flzy18001500226\u002FMoSim\u002Fblob\u002Fmain\u002FReferences\u002FAgent\u002FWorkflow\u002Fokwinds\u002Fmiscellany\u002Fagent\u002Fskills\u002Fprd-writing-guide\u002FSKILL.md",[],[],"2026-11-12 13:55:43"]