rrr
starsforks
How to install rrr?
dgk-dev/dgk-gpt/rrr
$npx -y skills add dgk-dev/dgk-gpt --skill rrrGLM-5 코드 리뷰. /rr의 상위 버전으로 더 강한 모델로 변경사항을 검토하고, Codex가 결과를 다시 검증해 유효한 이슈만 정리한다.
SKILL.md
| 1 | # /rrr |
| 2 | |
| 3 | Use `glm-review --model glm-5` as the deeper paid review pass, then validate the results against the code before you trust them. |
| 4 | |
| 5 | ## Default Flow |
| 6 | |
| 7 | 1. Determine the exact review scope first. |
| 8 | - commit hash known -> commit mode |
| 9 | - `staged` -> staged mode |
| 10 | - `pr` -> PR mode |
| 11 | - mixed unrelated local changes -> focused diff file |
| 12 | 2. Run a health check only when auth or connectivity looks questionable: |
| 13 | |
| 14 | ```bash |
| 15 | glm-review --model glm-5 --health |
| 16 | ``` |
| 17 | |
| 18 | 3. Run the review with the smallest correct input: |
| 19 | |
| 20 | ```bash |
| 21 | glm-review --model glm-5 |
| 22 | glm-review --model glm-5 --mode staged |
| 23 | glm-review --model glm-5 --mode pr |
| 24 | glm-review --model glm-5 --mode commit --ref <COMMIT_HASH> |
| 25 | glm-review --model glm-5 --diff-file /tmp/glm-review-diff.patch |
| 26 | ``` |
| 27 | |
| 28 | 4. Verify every reported issue in the actual code and diff. |
| 29 | 5. Keep only confirmed issues. |
| 30 | 6. If fixes are requested or clearly in scope, apply them and rerun the closest verification. |
| 31 | |
| 32 | ## Choosing Review Input |
| 33 | |
| 34 | - committed single change: |
| 35 | |
| 36 | ```bash |
| 37 | glm-review --model glm-5 --mode commit --ref <COMMIT_HASH> |
| 38 | ``` |
| 39 | |
| 40 | - committed subset of files: |
| 41 | |
| 42 | ```bash |
| 43 | glm-review --model glm-5 --mode commit --ref <COMMIT_HASH> --files src/a.ts src/b.ts |
| 44 | ``` |
| 45 | |
| 46 | - custom focused diff: |
| 47 | |
| 48 | ```bash |
| 49 | GIT_ROOT=$(git rev-parse --show-toplevel) |
| 50 | cd "$GIT_ROOT" && git diff HEAD -- <file1> <file2> ... > /tmp/glm-review-diff.patch |
| 51 | glm-review --model glm-5 --diff-file /tmp/glm-review-diff.patch |
| 52 | ``` |
| 53 | |
| 54 | If the diff is empty, stop and say there is nothing to review. |
| 55 | |
| 56 | ## Validation Rules |
| 57 | |
| 58 | - `glm-review` is a reviewer, not an oracle. |
| 59 | - Re-check referenced code paths before reporting a bug. |
| 60 | - Separate confirmed defects from debatable style feedback. |
| 61 | - If a finding depends on broader context, inspect the relevant surrounding code before accepting it. |
| 62 | |
| 63 | ## Error Handling |
| 64 | |
| 65 | - `command not found: glm-review` |
| 66 | Install it with `npm install -g glm-review` |
| 67 | - `ZAI_API_KEY not set` |
| 68 | Export `ZAI_API_KEY` in the shell or your normal secret-loading path |
| 69 | - auth or API failure |
| 70 | Re-run `glm-review --model glm-5 --health` after refreshing credentials |
| 71 | |
| 72 | ## Finish |
| 73 | |
| 74 | Return: |
| 75 | |
| 76 | - review scope used |
| 77 | - confirmed findings only |
| 78 | - notable false positives filtered out |
| 79 | - what verification ran after fixes, if any |
People who liked this also viewed
6 picksType
- find-skillsHelps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skSkillsJul 2026
- grill-meA relentless interview to sharpen a plan or design.SkillsJul 2026
- vercel-react-best-practicesVercel's official collection of agent skillsSkillsJul 2026
- improve-codebase-architectureScan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.SkillsJul 2026
- setup-matt-pocock-skillsConfigure this repo for the engineering skills — set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills.SkillsJul 2026
- triageMove issues and external PRs through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs.SkillsJul 2026
