PR Babysitter
openinterpreter/openinterpreter
Watches an open GitHub pull request until it merges, handling review comments, diagnosing CI failures and retrying flaky checks along the way.
Land explicitly requested changes in azu/irodr through a GitHub pull request and verified squash merge to master.
$ npx skills add azu/irodr --skill land -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install azu/irodr land --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/azu/irodr.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/land .claude/skills/land && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "land" agent skill from https://github.com/azu/irodr/tree/master/.agents/skills/land into .claude/skills/land/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "land", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/azu/irodr/tree/master/.agents/skills/landType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add azu/irodr --skill land -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install azu/irodr land --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/azu/irodr.git skills-src && mkdir -p .agents/skills && cp -r skills-src/.agents/skills/land .agents/skills/land && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "land" agent skill from https://github.com/azu/irodr/tree/master/.agents/skills/land into .agents/skills/land/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "land", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add azu/irodr --skill land -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install azu/irodr land --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/azu/irodr.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/.agents/skills/land .cursor/skills/land && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "land" agent skill from https://github.com/azu/irodr/tree/master/.agents/skills/land into .cursor/skills/land/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "land", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/azu/irodr.git --path .agents/skills/land--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add azu/irodr --skill land -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install azu/irodr land --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/azu/irodr.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/.agents/skills/land .gemini/skills/land && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "land" agent skill from https://github.com/azu/irodr/tree/master/.agents/skills/land into .gemini/skills/land/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "land", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install azu/irodr landInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add azu/irodr --skill land -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/azu/irodr.git skills-src && mkdir -p .github/skills && cp -r skills-src/.agents/skills/land .github/skills/land && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "land" agent skill from https://github.com/azu/irodr/tree/master/.agents/skills/land into .github/skills/land/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "land", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add azu/irodr --skill land -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install azu/irodr land --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/azu/irodr.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/.agents/skills/land .opencode/skills/land && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "land" agent skill from https://github.com/azu/irodr/tree/master/.agents/skills/land into .opencode/skills/land/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "land", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
landLand explicitly requested changes in azu/irodr through a GitHub pull request and verified squash merge to master.
Land is an agent skill from azu/irodr. Land explicitly requested changes in azu/irodr through a GitHub pull request and verified squash merge to master. Invoke only when the user requests landing or merging, including Land Changes or /land, not merely for review, preparation, passing checks, or installing this skill.
Its SKILL.md is about 3.5k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.
It works with GitHub. The repository describes itself as: RSS reader client like LDR for Inoreader. The licence is MIT.
4 steps, taken from the first numbered list in SKILL.md.
Read from SKILL.md and the folder at commit e589bbc. It shows what the files ask for, not the result of running them.
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Shell commands in SKILL.md call:
ghgitpnpmmisenodeFrom the folder's file list and the shell code blocks in SKILL.md.
Hosts in commands or code, which the agent is likely to contact:
github.comFrom URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Land loads about 3.5k tokens when it runs. Until then it costs about 71 tokens; SKILL.md has 1,803 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
The full file from azu/irodr at commit e589bbc, republished under its MIT licence (© azu). 1,803 words, ~3,532 tokens.
.claude/skills/land/SKILL.md (or your agent's skills folder).This skill applies only to azu/irodr. An explicit landing request authorizes
preparation, commits, topic-branch publication, PR creation, and merging through
the workflow below. /land and accepting an offer to run this installed skill
are explicit landing requests. Proceed without asking for the same permission
again. Installing this skill alone does not authorize execution.
Use the attached Delta worktree, not the user's primary checkout. Land only the requested change; preserve unrelated staged, unstaged, and untracked work. Never publish secrets, disable signing or hooks, force-push, rewrite shared history, bypass protection, or change repository settings to make landing pass. If scope is genuinely ambiguous, ask one focused question before proceeding.
When the user requests landing this skill first and another change afterward, land them as separate PRs in that order. Stage only the skill for the first PR; keep the other changes out of its commits. Carry the remaining work forward without discarding it, and create the second PR from the updated target.
Read applicable AGENT.md/AGENTS.md, current contribution instructions,
templates, and the relevant change. Inspect:
git --no-optional-locks status --short --branch
git diff
git diff --cached
git remote -v
git log -8 --onelineIdentify existing relevant commits as well as uncommitted work. Do not stage all files blindly or replace the user's partial staging. Check for unresolved merge markers and Git operations before starting another operation.
Verify the publication remote is GitHub azu/irodr; normally it is origin.
Never push to Delta's local backlink. Verify the default branch, currently
master, and the intended PR base. If the destination differs from this
scope, stop and clarify instead of silently choosing main or another repo.
Check gh auth status without displaying credentials and confirm Git push
access. Authentication and tools can differ between machines.
Inspect destination settings using gh api / gh pr view:
repos/azu/irodr/branches/master and
repos/azu/irodr/rules/branches/master.During setup, master was unprotected, the applicable-rules endpoint returned
no rules, and squash merges were allowed. These are observations, not permanent
exemptions. The token could not read the branch-protection administration
endpoint: do not interpret a 403 as "no requirements." An explicit unprotected
branch response plus successfully retrieved empty applicable rules can establish
that no branch rules apply. Otherwise, if requirements cannot be established,
stop and explain the missing access. Never use gh pr merge --admin.
Follow the contribution policy in README.md ("Contributing"): a feature
branch and pull request, rather than a direct target-branch push. Use a fork
if upstream topic-branch push permission is unavailable; stop if no authorized
publication destination is available. Respect CODE_OF_CONDUCT.md. For a
newly discovered nonpublic vulnerability, follow the inherited security policy
at https://github.com/azu/.github/blob/master/SECURITY.md before publishing
sensitive details.
No additional CLA, mandatory human-authored submission text, changelog entry, or fixed review count was found during setup. Recheck current policies and templates; enforce any applicable unmet requirements, including their conditions and exceptions. Do not invent requirements or ask again about requirements already satisfied. Describe behavior accurately in affected documentation; do not claim tests that were not run.
Use the Node version in .node-version (24.14.1 at setup) and pnpm from
package.json#packageManager (10.34.5 at setup). CI reads .node-version through
voidzero-dev/setup-vp; local verification does not replace CI. Do not silently use
whichever Node happens to be first on PATH.
Let the runtime manager read the project version files rather than overriding
the version in each command. For mise, the existing
settings.idiomatic_version_file_enable_tools setting should include "node"
so it reads .node-version automatically. Verify the active versions:
node --version
pnpm --versionUse the ordinary package commands below when those versions match. If the
non-interactive terminal has not activated mise, prefix commands with
mise exec -- to use its project-selected versions, without an explicit
node@... override. Sources: .node-version and package.json#packageManager;
the optional wrapper syntax was checked against mise exec --help.
Ask before installing global tooling or changing the user's external configuration.
Fetch the verified publication remote. Reuse the relevant topic branch/PR
where appropriate; otherwise create a descriptive feature/... branch,
including when the work started as uncommitted edits on master. Check the
entire base-to-head diff so old or unrelated commits do not slip into the PR.
Do not discard existing work or stash it without a clear recovery plan.
Install project dependencies when needed:
pnpm install --frozen-lockfileSource: .github/workflows/test.yml installs with vp install --frozen-lockfile
(Vite+ runs the pnpm version from package.json#packageManager);
package.json#packageManager and pnpm-lock.yaml define the dependency state.
--frozen-lockfile prevents incidental lockfile changes.
package.json#scripts.prepare runs vp config --no-agent, which installs the
Vite+ hook dispatcher for .vite-hooks. Preserve that hook behavior. If the
manifest and lockfile are inconsistent, fix only the intended dependency
change and rerun verification; do not bypass the failure.
For code, dependency, or build/test configuration changes, run:
pnpm run check
pnpm test
pnpm run build
CI=true pnpm run test:e2eSources: .github/workflows/test.yml runs vp check, vp test and vp build
(job "Check, unit test and build") and vp run test:e2e (job "Integration tests").
package.json#scripts.check runs vp check (Oxfmt, Oxlint with type checking);
package.json#scripts.test runs the Vitest unit tests once (vp test does not watch);
package.json#scripts.test:e2e builds with --mode e2e and runs Playwright against
the fake APIs in e2e/fake-api. CI=true makes Playwright start fresh servers.
Playwright needs Chromium (pnpm exec playwright install chromium, or set
PLAYWRIGHT_CHROMIUM_EXECUTABLE).
For skill/documentation-only changes, local verification can instead check
syntax, links, command references, frontmatter, and the scoped diff. This
exception is local only: required remote checks still must pass.
Check git diff --check for every change.
If local verification fails due to an environment issue, use the declared runtime and retry. Report genuine unavailable prerequisites or failed checks; do not label them passed, replace tests with mocks, or weaken the checks.
Stage only the intended paths/hunks. Use an English conventional commit
subject (feat:, fix:, docs:, chore:, etc.), following existing history.
Use a non-interactive commit:
GIT_EDITOR=true git commit -m "type: concise description"Preserve the user's signing configuration (SSH commit signing was enabled at
setup). Never use --no-verify to bypass .vite-hooks/pre-commit, which runs
vp staged (the staged block in vite.config.ts formats and lints staged files)
and vp test.
Inspect the resulting commit and any hook changes. Verification must cover
the final contents, not the pre-hook or earlier version.
Push the topic branch to the verified publication remote, never to local or
directly to master. Use ordinary non-force pushes. On rejection, inspect the
cause; do not bypass permissions through API-based Git writes.
Reuse an existing PR with matching scope/base, or create a ready-for-review
PR targeting master using gh pr create with explicit repository, base,
head, title, and body. Pass text using non-interactive options; if using a
body file, write it under the current scratch directory. Follow any template.
Include the problem, actual behavior, verification results, and relevant issue
links without inventing issue requirements. A draft PR must become ready and
satisfy its review requirements before merging.
Refresh the target. When integration is needed, merge the latest target into
the topic branch using GIT_EDITOR=true git merge origin/master (substitute
the verified remote if different). Do not rebase published history.
Conflict preference: resolve automatically when intent is clear. Preserve both sides' intended behavior and unrelated work. For ambiguous behavior, unsafe changes, or uncertain ownership, stop and ask. Do not blindly choose ours/theirs or delete a lockfile to force resolution. After any resolution or other code change, repeat applicable local verification, commit non-interactively, push normally, and obtain new CI/review results.
Verify all applicable required checks and reviews for the exact current head.
Always require the repository's test workflow, including every current
job ("Check, unit test and build" and "Integration tests"), even if branch
protection does not mark it required. Inspect both workflow runs and PR check/status results.
Check a Netlify deploy preview when present; do not mistake neutral informational
Netlify checks for a successful build.
gh pr checks PR --repo azu/irodr --watch --interval 10 can wait for checks;
bound each watch with the terminal timeout and re-query on timeout. A successful
CLI exit or empty list is not enough: confirm the expected workflow actually
ran for this head/integration, completed, and passed. Use gh run view and
check details as needed to verify SHA, event, conclusion, and result URL.
Pending, failing, cancelled, missing, stale, or unverifiable required checks
block landing. Informational skipped/neutral checks do not substitute for the
required tests. Do not override requested changes or required reviews.
Immediately before merging, re-read head SHA, target SHA, mergeability, required reviews and checks. If either relevant SHA changed, reassess and obtain verification for the new changes/integration.
Squash merge with an exact-head guard:
gh pr merge PR --repo azu/irodr --squash --match-head-commit VERIFIED_HEAD_SHAReplace the uppercase arguments with verified values. Flags were checked
against gh pr merge --help. This is the approved workflow, not a claim that
squash is mandated by repository policy. If squash is no longer permitted,
stop and ask rather than silently changing strategy. If a merge queue becomes
required, honor it without --admin and wait for the actual merge and all
required queue checks; entering a queue is not completion.
Confirm the PR is MERGED into the intended repository and master, retrieve
its actual merge commit SHA, fetch the destination, and verify that commit is
reachable from the remote target. Account for squash commits having a different
SHA from the topic branch.
Confirm the target-branch test workflow passed for the landed commit, not
merely for a prior PR revision. Source: .github/workflows/test.yml runs on both
push and pull_request. If post-merge CI fails or cannot be verified, explicitly
say the change has landed but destination verification failed; do not pretend it
is unmerged or automatically revert shared history.
Do not reset the user's checkout, discard remaining work, or delete branches containing other work. Leave the topic branch available unless cleanup was requested. Finish by checking working-tree state and explaining any remaining changes. A prepared commit, branch push, open PR, queued merge, or passing local build is not successful landing.
When running in a subthread with report_subthread_status available, report the
verified outcome to the parent; otherwise report directly in this conversation.
Use success only after destination and CI verification. Use failure for a
failed attempt or genuine blocker, describing whether landing occurred.
Continue safe recovery when permitted and report an updated result after
verification. Do not use this tool for skill installation or routine progress.
Keep the title to a few sentence-case words and the description to one short line. Link the short commit SHA and actual CI result using verified URLs; omit unavailable links rather than constructing fictional ones. Examples of wording:
Landed on master — linked short SHA, then linked CI passed.Blocked by CI — linked failed run and commit, then
Not landed.Push blocked — state the required access and that the
change has not landed.Post-merge CI failed — linked landed SHA and failed run,
explicitly saying it is already on master.Ask necessary questions in the conversation, not in a status event.
© azu, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Just SKILL.md in .agents/skills/land of azu/irodr.
Open the folder on GitHubat commit e589bbc
Land next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Land this skillazu/irodr | 141 | — | ~3.5k | Automated safety check: Pass | MIT | |
| PR Babysitteropeninterpreter/openinterpreter | 69k | 3 repos | ~4.2k | Automated safety check: Pass | Apache-2.0 | |
| Greplooponyx-dot-app/onyx | 32k | 4 repos | ~3.3k | Automated safety check: Pass | MIT | |
| GitHub Deep Researchbytedance/deer-flow | 84k | 4 repos | ~1.3k | Automated safety check: Pass | MIT | |
| Diagnosing Superpowers Sessionsobra/superpowers | 297k | 3 repos | ~1.7k | Automated safety check: Pass | MIT | |
| Update V8 Versionopeninterpreter/openinterpreter | 69k | 2 repos | ~845 | Automated safety check: Pass | Apache-2.0 |
openinterpreter/openinterpreter
Watches an open GitHub pull request until it merges, handling review comments, diagnosing CI failures and retrying flaky checks along the way.
onyx-dot-app/onyx
Iteratively improves a PR (GitHub), MR (GitLab), or shelved changelist (Perforce) until Greptile gives it a 5/5 confidence score with zero unresolved comments.
bytedance/deer-flow
Researches a GitHub repository over four rounds using the GitHub API and web search, then writes a structured markdown report with timeline, metrics and Mermaid diagrams.
obra/superpowers
Investigates a session where Superpowers went wrong, reads the transcripts on disk and produces an evidence-cited report, optionally prepared as a bug report for the maintainers.
openinterpreter/openinterpreter
Bumps the pinned v8 and rusty_v8 versions in Codex, validates the release-candidate path with the v8-canary check, and traces failures to upstream build changes.
mvanhorn/last30days-skill
Research what people actually say about any topic in the last 30 days.
Works with
Land explicitly requested changes in azu/irodr through a GitHub pull request and verified squash merge to master. Land is an agent skill from azu/irodr. Land explicitly requested changes in azu/irodr through a GitHub pull request and verified squash merge to master.
Land fits situations like: requests landing; including Land Changes; not merely for review; installing this skill.
Run `npx skills add azu/irodr --skill land -a claude-code`. Or copy the skill folder (.agents/skills/land in azu/irodr) into .claude/skills/land in your project. Claude Code loads it when a task matches its description.
Run `npx skills add azu/irodr --skill land -a codex`. Or copy the skill folder (.agents/skills/land in azu/irodr) into .agents/skills/land in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add azu/irodr --skill land -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/land, .gemini/skills/land, .github/skills/land and .opencode/skills/land in your project.
Going by SKILL.md and its folder, Land needs the command-line tools its instructions call (gh, git, pnpm, mise and node).
SKILL.md names 1 domain. In commands or code: github.com; the agent is likely to contact it when it follows the instructions. This is read from the text; nothing was executed.
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
Land is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
About 3.5k tokens (SKILL.md is roughly 14k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
Skills that share tags, products or a category with Land: PR Babysitter (openinterpreter/openinterpreter, 69k stars), Greploop (onyx-dot-app/onyx, 32k stars), GitHub Deep Research (bytedance/deer-flow, 84k stars) and Diagnosing Superpowers Sessions (obra/superpowers, 297k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
azu (a GitHub user) maintains it in azu/irodr, which has 141 GitHub stars. The repository was last updated on October 10, 2026.
Source: azu/irodr on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.