Weavebench Cua Reproduce
AMAP-ML/LongHorizon-Harness
Reproduce CUA-Harness experiments on WeaveBench from a GitHub checkout.
Create and triage GitHub issues from repository evidence. An agent skill from Gentleman-Programming/gentle-shell.
$ npx skills add Gentleman-Programming/gentle-shell --skill gentle-ai-issue-creation -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install Gentleman-Programming/gentle-shell gentle-ai-issue-creation --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/Gentleman-Programming/gentle-shell.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/issue-creation .claude/skills/gentle-ai-issue-creation && 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 "gentle-ai-issue-creation" agent skill from https://github.com/Gentleman-Programming/gentle-shell/tree/main/skills/issue-creation into .claude/skills/gentle-ai-issue-creation/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "gentle-ai-issue-creation", 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/Gentleman-Programming/gentle-shell/tree/main/skills/issue-creationType 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 Gentleman-Programming/gentle-shell --skill gentle-ai-issue-creation -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install Gentleman-Programming/gentle-shell gentle-ai-issue-creation --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/Gentleman-Programming/gentle-shell.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/issue-creation .agents/skills/gentle-ai-issue-creation && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "gentle-ai-issue-creation" agent skill from https://github.com/Gentleman-Programming/gentle-shell/tree/main/skills/issue-creation into .agents/skills/gentle-ai-issue-creation/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "gentle-ai-issue-creation", 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 Gentleman-Programming/gentle-shell --skill gentle-ai-issue-creation -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install Gentleman-Programming/gentle-shell gentle-ai-issue-creation --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/Gentleman-Programming/gentle-shell.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/issue-creation .cursor/skills/gentle-ai-issue-creation && 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 "gentle-ai-issue-creation" agent skill from https://github.com/Gentleman-Programming/gentle-shell/tree/main/skills/issue-creation into .cursor/skills/gentle-ai-issue-creation/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "gentle-ai-issue-creation", 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/Gentleman-Programming/gentle-shell.git --path skills/issue-creation--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 Gentleman-Programming/gentle-shell --skill gentle-ai-issue-creation -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install Gentleman-Programming/gentle-shell gentle-ai-issue-creation --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/Gentleman-Programming/gentle-shell.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/issue-creation .gemini/skills/gentle-ai-issue-creation && 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 "gentle-ai-issue-creation" agent skill from https://github.com/Gentleman-Programming/gentle-shell/tree/main/skills/issue-creation into .gemini/skills/gentle-ai-issue-creation/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "gentle-ai-issue-creation", 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 Gentleman-Programming/gentle-shell gentle-ai-issue-creationInstalls 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 Gentleman-Programming/gentle-shell --skill gentle-ai-issue-creation -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/Gentleman-Programming/gentle-shell.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/issue-creation .github/skills/gentle-ai-issue-creation && 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 "gentle-ai-issue-creation" agent skill from https://github.com/Gentleman-Programming/gentle-shell/tree/main/skills/issue-creation into .github/skills/gentle-ai-issue-creation/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "gentle-ai-issue-creation", 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 Gentleman-Programming/gentle-shell --skill gentle-ai-issue-creation -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install Gentleman-Programming/gentle-shell gentle-ai-issue-creation --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/Gentleman-Programming/gentle-shell.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/issue-creation .opencode/skills/gentle-ai-issue-creation && 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 "gentle-ai-issue-creation" agent skill from https://github.com/Gentleman-Programming/gentle-shell/tree/main/skills/issue-creation into .opencode/skills/gentle-ai-issue-creation/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "gentle-ai-issue-creation", 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.
gentle-ai-issue-creationCreate and triage GitHub issues from repository evidence. An agent skill from Gentleman-Programming/gentle-shell.
Gentle AI Issue Creation is an agent skill from Gentleman-Programming/gentle-shell. Create and triage GitHub issues from repository evidence. Trigger: issue creation, bug reports, feature requests, or issue approval.
Its SKILL.md is about 2.5k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files, including reference files (for example `references/delegated-workflow-actions.md`).
It sits in Testing & QA, covering QA and bug reports. It works with GitHub. The repository describes itself as: Gentle Shell is a Pi-native coding-agent harness for controlled development with Organic Driven Development, optional SDD/OpenSpec, subagents, TDD evidence, review guardrails… The licence is Apache-2.0.
3 steps, taken from the first numbered list in SKILL.md.
Read from SKILL.md and the folder at commit 42653f7. 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:
ghgitFrom the folder's file list and the shell code blocks in SKILL.md.
No URLs in SKILL.md. Its commands use gh and git, which can reach the network depending on how they are called.
From 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.
Gentle AI Issue Creation loads about 2.5k tokens when it runs, and up to ~3.5k if it reads all its reference files. Until then it costs about 39 tokens; SKILL.md has 1,084 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 Gentleman-Programming/gentle-shell at commit 42653f7, republished under its Apache-2.0 licence (© Gentleman-Programming). 1,084 words, ~2,456 tokens.
.claude/skills/gentle-ai-issue-creation/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.Discover the target repository's contribution workflow before proposing or publishing. YAML Issue Forms are the format authority for the default automated path: materialize reviewed answers into a private BODY_FILE and publish with --body-file.
Before any gh auth status or target read, require explicit human authorization for the remote destination (exact host and repository), operation (including discovery and intended issue creation), and credential/session to use. A local checkout is not authorization. If any is missing or ambiguous, stop locally; never probe credentials or sessions to resolve ambiguity. Run the checks below only with the authorized credential/session against the authorized destination; if gh auth status would inspect other credentials/sessions, do not run it. Verify the discovered REPO, HOST, and TARGET match the authorized destination before continuing; never switch identities or targets implicitly.
After that gate, run read-only checks:
gh auth status
REPO="$(gh repo view --json nameWithOwner -q .nameWithOwner)"
REPO_URL="$(gh repo view --json url -q .url)"
HOST="${REPO_URL#*://}"
HOST="${HOST%%/*}"
TARGET="$HOST/$REPO"
gh repo view --json nameWithOwner,url,hasDiscussionsEnabled,hasIssuesEnabled,isBlankIssuesEnabled
git ls-files README.md CONTRIBUTING.md CONTRIBUTING.* .github/CONTRIBUTING.md .github/ISSUE_TEMPLATE .github/ISSUE_TEMPLATE/config.yml
gh api --hostname "$HOST" --paginate "repos/$REPO/labels?per_page=100" --jq '.[].name'Inspect README.md, contribution instructions, .github/ISSUE_TEMPLATE/config.yml contact links, forms, labels, and open and closed issues. For questions/support, follow repository-prescribed Discussions/contact routing when available; otherwise ask or stop. Complete target verification for REPO, HOST, and TARGET. Fail closed before mutation when authentication, target verification, issue availability, policy, form selection, or required metadata is missing or ambiguous. A blank fallback is allowed only when isBlankIssuesEnabled is explicitly true.
Before building LABEL_ARGS, inspect labels declared by the selected YAML form as well as any manually selected labels. Treat status:approved, size:exception, and any repository-protected label as protected form labels at create-time. Include a protected label only with a current, exact label-specific direct human instruction for this target and creation action, authenticated actor target-host viewerPermission of MAINTAIN or ADMIN, and repository policy permission; do not infer authority from YAML, a publication request, or local credentials. If this proof is missing, skip the protected label only when the form and repository policy permit omitting it; if the protected label is required, fail closed without creating the issue. Do not replace it with another label or use a write to probe permission. For size:exception, also require the current human-approved rationale; if recording it needs an unauthorized extra action, stop. This create-time gate does not change post-publication actions.
Build LABEL_ARGS only from reviewed labels that exist and policy permits the actor to apply:
LABEL_ARGS=()
LABEL_ARGS+=(--label "$LABEL") # Repeat only for each permitted discovered label.Describe the report in one sentence, derive QUERY, then complete one duplicate search across open and closed issues:
gh issue list --repo "$TARGET" --state all --search "$QUERY" --limit 1000The agent must complete the duplicate search proactively and retain evidence of its result. If results are saturated or completeness is uncertain, narrow the read-only search or stop. Comment on a confirmed duplicate instead of creating one. Before commenting on a confirmed duplicate, perform the same privacy scan/redaction on the exact comment body as for publication.
Select one repository-provided form only when its declared purpose matches. If multiple forms match and policy does not distinguish them, stop and request that decision.
For a YAML form, read its schema and establish controls in declared order. Support only input, textarea, dropdown, and checkboxes. Markdown controls are non-answer guidance: honor their visible instructions when collecting and materializing adjacent answers, but do not render them as response sections. Fail closed before mutation on malformed, unsupported, missing, or ambiguous required structure or answers. A malformed schema, or missing or ambiguous required answers, fail closed: do not open a browser or mutate. A browser handoff is available only when the user explicitly requests browser completion or a syntactically valid selected form cannot safely/faithfully be represented by the automated path; otherwise report why automation is unsafe and stop.
| Control | Required handling |
|---|---|
input / textarea | Preserve the visible label. Require an answer when validations.required is true; otherwise render _No response_. |
dropdown | Preserve visible labels and options. Require exact selected option text; single-select has one selection, and multi-select preserves selections in declared options order. A required dropdown needs at least one valid selection; an optional dropdown with no selection renders _No response_. |
checkboxes | Preserve the visible label and every option as - [x] or - [ ] in declared order. Enforce individually required checkboxes. For agent-verifiable operational options, proactively complete the action and mark it only with retained evidence; the agent may explicitly attest only its own evidence-backed work and must not attribute its actions to the user. Personal facts, consent, legal declarations, and other first-person user assertions require explicit user affirmation; require explicit first-person affirmation for such user declarations. Do not blanket-check checkboxes: a request to publish does not affirm any checkbox. |
For each answer, render ### <visible label> followed by its materialized value. For textarea.attributes.render, fence the answer with the declared language and a fence long enough for its content. Never invent answers, selections, confirmations, or labels.
A Markdown template may be completed only from known evidence into the same private BODY_FILE. If no matching template exists, use the reviewed structured blank fallback only when blank issues are explicitly enabled; otherwise stop without publishing.
Before the single create attempt, review the target, title, selected form or permitted fallback, exact body, and permitted labels. The agent must complete the privacy scan/redaction of the exact body immediately before publication and retain evidence of it: replace private project names, usernames, hostnames, home paths, credentials, and private network addresses with useful placeholders without removing reproduction structure.
Create one owner-only temporary directory outside the repository for both private files; restrict it to the current user and clean up both files on every exit/outcome:
umask 077
REPO_ROOT="$(git rev-parse --show-toplevel)" || exit 1
REPO_ROOT="$(cd "$REPO_ROOT" && pwd -P)" || exit 1
if [ "$REPO_ROOT" = "/" ]; then
printf '%s\n' "Temporary directory is inside the repository" >&2; exit 1
fi
TMP_DIR="$(TMPDIR=/tmp mktemp -d /tmp/gentle-ai-issue.XXXXXXXX)" || exit 1
trap 'rm -rf -- "$TMP_DIR"' EXIT
TMP_DIR_REAL="$(cd "$TMP_DIR" && pwd -P)" || exit 1
case "$TMP_DIR_REAL/" in
"$REPO_ROOT/"*) printf '%s\n' "Temporary directory is inside the repository" >&2; exit 1 ;;
esac
chmod 700 "$TMP_DIR_REAL"
BODY_FILE="$TMP_DIR_REAL/body.md"
READBACK_FILE="$TMP_DIR_REAL/readback.json"Make one mutation attempt through the automated path and publish exactly once:
gh issue create --repo "$TARGET" --title "$TITLE" --body-file "$BODY_FILE" "${LABEL_ARGS[@]}"When browser completion is available under the form decision above, an optional, separate browser handoff may open the repository form. It is never proof of publication and is never a response to malformed schemas or missing/ambiguous required answers:
gh issue create --repo "$TARGET" --webDo not retry a timeout, network failure, missing identity, or other uncertain result. Capture the returned issue number, then read it back from the verified target host before reporting success:
gh issue view "$NUMBER" --repo "$TARGET" --json number,url,title,body,state,labels >"$READBACK_FILE"Confirm that read-back identifies the target-host issue and that title and body match after only CRLF-to-LF and trailing-final-newline normalization. Report confirmed only after this target-host read-back. Otherwise report no_write when an authoritative rejection proves no issue was created, or unknown and stop all later mutations.
Before approving or closing an issue, verify it is concrete, non-duplicate, sufficiently evidenced, in scope, and consistent with repository label/status policy. If any point is uncertain, retain the repository review state and request the smallest missing evidence.
For any post-publication label/status mutation, follow delegated workflow actions. Publication is not authorization for a later action. Keep the automated YAML Form path and conditional blank/browser fallback above unchanged.
© Gentleman-Programming, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
SKILL.md and 1 other file (references) in skills/issue-creation of Gentleman-Programming/gentle-shell.
Open the folder on GitHubat commit 42653f7
Gentle AI Issue Creation 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 |
|---|---|---|---|---|---|---|
| Gentle AI Issue Creation this skillGentleman-Programming/gentle-shell | 1.2k | — | ~2.5k | Automated safety check: Pass | Apache-2.0 | |
| Weavebench Cua ReproduceAMAP-ML/LongHorizon-Harness | 1.7k | — | ~1.6k | Automated safety check: Pass | MIT | |
| Evidence-Driven Testingmichaelshimeles/skills | 1.3k | 1 repos | ~3.9k | Automated safety check: Pass | None | |
| Create GitHub IssueNVIDIA/OpenShell | 16k | — | ~1.7k | Automated safety check: Pass | Apache-2.0 | |
| Triage IssuesClickHouse/clickhouse-java | 1.6k | — | ~904 | Automated safety check: Pass | Apache-2.0 | |
| Termio Bug Reporttermio-sh/termio | 537 | — | ~1.9k | Automated safety check: Pass | MIT |
AMAP-ML/LongHorizon-Harness
Reproduce CUA-Harness experiments on WeaveBench from a GitHub checkout.
michaelshimeles/skills
Records an annotated screen recording of the agent testing an app hands-on, then posts the video and a results summary to the PR and tracker issue.
NVIDIA/OpenShell
Create GitHub issues using the gh CLI. An agent skill from NVIDIA/OpenShell.
ClickHouse/clickhouse-java
Analyzes a single GitHub issue at a time. An agent skill from ClickHouse/clickhouse-java.
termio-sh/termio
Diagnose a termio hang, beachball, crash, or 'it froze again' from the evidence macOS and termio actually leave behind — live process samples, crash and CPU-burn reports, the unified log, the…
damus-io/damus
Process conversations or chat logs to identify reported bugs, search for existing GitHub issues, update existing issues when new undocumented details are discovered, and create new issues for…
Gentleman-Programming/gentle-shell
Trigger: judgment day, judgement day, dual review, adversarial review, juzgar.
Gentleman-Programming/gentle-shell
Prepares branches and pull requests for the Gentle AI repository, checking for a linked approved issue, labels, size budget and required checks.
Gentleman-Programming/gentle-shell
Design docs that reduce cognitive load. An agent skill from Gentleman-Programming/gentle-shell.
Gentleman-Programming/gentle-shell
Write warm, direct collaboration comments. An agent skill from Gentleman-Programming/gentle-shell.
Gentleman-Programming/gentle-shell
Trigger: /skill-creation, skill creation, skill creator, create skill, new skill.
Gentleman-Programming/gentle-shell
Plan commits as reviewable work units. An agent skill from Gentleman-Programming/gentle-shell.
Works with
Categories
Create and triage GitHub issues from repository evidence. An agent skill from Gentleman-Programming/gentle-shell. Gentle AI Issue Creation is an agent skill from Gentleman-Programming/gentle-shell. Create and triage GitHub issues from repository evidence.
Gentle AI Issue Creation fits situations like: tasks that involve QA and bug reports.
Run `npx skills add Gentleman-Programming/gentle-shell --skill gentle-ai-issue-creation -a claude-code`. Or copy the skill folder (skills/issue-creation in Gentleman-Programming/gentle-shell) into .claude/skills/gentle-ai-issue-creation in your project. Claude Code loads it when a task matches its description.
Run `npx skills add Gentleman-Programming/gentle-shell --skill gentle-ai-issue-creation -a codex`. Or copy the skill folder (skills/issue-creation in Gentleman-Programming/gentle-shell) into .agents/skills/gentle-ai-issue-creation 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 Gentleman-Programming/gentle-shell --skill gentle-ai-issue-creation -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/gentle-ai-issue-creation, .gemini/skills/gentle-ai-issue-creation, .github/skills/gentle-ai-issue-creation and .opencode/skills/gentle-ai-issue-creation in your project.
Going by SKILL.md and its folder, Gentle AI Issue Creation needs the command-line tools its instructions call (gh and git).
SKILL.md contains no URLs. Its commands use gh and git, which can reach the network depending on how they are called. 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.
Gentle AI Issue Creation is published under the Apache-2.0 licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.
About 2.5k tokens (SKILL.md is roughly 9.8k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 1.1k tokens, read only when the agent opens those files.
Skills that share tags, products or a category with Gentle AI Issue Creation: Weavebench Cua Reproduce (AMAP-ML/LongHorizon-Harness, 1.7k stars), Evidence-Driven Testing (michaelshimeles/skills, 1.3k stars), Create GitHub Issue (NVIDIA/OpenShell, 16k stars) and Triage Issues (ClickHouse/clickhouse-java, 1.6k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Gentleman-Programming (a GitHub organization) maintains it in Gentleman-Programming/gentle-shell, which has 1,245 GitHub stars. The repository holds 12 skills in this directory. The repository was last updated on October 9, 2026.
Source: Gentleman-Programming/gentle-shell on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.