gh-ghent

gh-ghent — Agentic PR Monitoring

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "gh-ghent" with this command: npx skills add indrasvat/gh-ghent/indrasvat-gh-ghent-gh-ghent

gh-ghent — Agentic PR Monitoring

gh-ghent is a GitHub CLI extension that gives you structured access to PR review state: unresolved threads, CI check status, annotations, logs, and merge readiness.

Prerequisite: gh extension install indrasvat/gh-ghent

When to Use

  • Check PR merge readiness in a single call

  • Diagnose CI failures with error log excerpts and annotations

  • Read unresolved review comments with file:line locations

  • Resolve review threads after fixing the requested changes

  • Reply to review threads to acknowledge feedback

  • Monitor CI in a polling loop until checks complete

Agent Mode

Always use these flags for structured, parseable output:

gh ghent <command> --pr <N> --format json --no-tui

  • --no-tui forces pipe mode even in pseudo-TTY environments

  • --format json returns structured JSON (default in pipe mode)

  • --format xml returns tagged XML with semantic attributes

  • --format md returns readable Markdown

Recommended: Start with Summary

For most agent tasks, start with gh ghent summary --pr N --logs --format json --no-tui . This single call returns: unresolved threads with file:line, CI check status with annotations and error log excerpts, approval state, and merge readiness. Only use individual commands (comments , checks ) when you need deeper detail or specific actions (resolve , reply ).

Quick Start

  1. Get full PR status with failure diagnostics

gh ghent summary --pr 42 --logs --format json --no-tui

Returns: unresolved threads, CI checks with annotations and log excerpts, approvals, is_merge_ready . Exit code 0 = merge-ready, 1 = not ready.

  1. Wait for CI, then get full report

gh ghent summary --pr 42 --watch --format json --no-tui

Polls CI until terminal status (stderr), then outputs full summary (stdout). Pipe-friendly: gh ghent summary --pr 42 --watch --format json 2>/dev/null | jq

  1. Quick merge-readiness gate

gh ghent summary --pr 42 --quiet

Silent exit 0 if merge-ready, full output + exit 1 if not ready.

  1. Drill down when needed

gh ghent comments --pr 42 --format json --no-tui # Detailed threads gh ghent checks --pr 42 --logs --format json --no-tui # Detailed check runs gh ghent resolve --pr 42 --all # Batch resolve gh ghent reply --pr 42 --thread PRRT_... --body "Fixed" # Reply to thread

Commands

Command Purpose Key Flags

summary

Full PR status + failure diagnostics --logs , --watch , --quiet , --compact

comments

Unresolved review threads --group-by file|author|status

checks

CI status + annotations --logs , --watch

resolve

Resolve/unresolve threads --thread , --all , --file , --author , --dry-run , --unresolve

reply

Reply to a thread --thread , --body , --body-file

For complete flag reference, see references/command-reference.md.

Global Flags

Flag Short Default Description

--pr

(required) PR number

--repo

-R

current repo Repository in OWNER/REPO format

--format

-f

json

Output format: json , md , xml

--no-tui

false

Force pipe mode (for agents in pseudo-TTY)

--since

Filter by time: ISO 8601 or relative (1h , 30m , 2d , 1w )

--verbose

false

Include additional context

--debug

false

Debug logging to stderr

Troubleshooting: If output seems incomplete (e.g., empty log_excerpt , missing checks), add --debug to see API calls and timing on stderr. This reveals 404s from expired logs, rate limits, and context cancellations without changing stdout.

Core Workflow

1. Get everything in one call

SUMMARY=$(gh ghent summary --pr 42 --logs --format json --no-tui)

2. Check merge readiness

echo "$SUMMARY" | jq '.is_merge_ready'

3. If CI failing, read the log excerpts (covers failure, timed_out, cancelled, etc.)

echo "$SUMMARY" | jq '.checks.checks[] | select(.log_excerpt) | {name, conclusion, log_excerpt, annotations}'

4. If threads need fixing, read them

echo "$SUMMARY" | jq '.comments.threads[] | {path, line, body: .comments[0].body}'

5. Fix code, then resolve + reply

Thread IDs are in comments output at .threads[].id (e.g., PRRT_kwDO...)

gh ghent resolve --pr 42 --all

Exit Codes

Exit codes let you branch logic without parsing output:

Command 0 1 2 3

summary

merge-ready not ready error —

comments

no unresolved has unresolved error —

checks

all pass failure error pending

resolve

all success partial failure total failure —

reply

posted thread not found error —

Error (exit 2): authentication failure, rate limit, or resource not found.

For detailed exit code usage patterns, see references/exit-codes.md.

Incremental Monitoring with --since

Use --since to see only what changed since your last check:

Only threads with activity in the last hour

gh ghent comments --pr 42 --since 1h --format json --no-tui

Only checks completed in the last 30 minutes

gh ghent checks --pr 42 --since 30m --format json --no-tui

Supported formats: 1h , 30m , 2d , 1w , 24h , or ISO 8601 (2026-02-23T00:00:00Z ).

Compact Summary for Minimal Context

gh ghent summary --pr 42 --compact --logs --format json --no-tui

Returns a flat object with pr_age , last_update , review_cycles , is_merge_ready , unresolved count, pass_count , fail_count , a threads digest, and failed_checks

with annotations and log excerpts. Uses minimal tokens — ideal for agent polling loops.

Group Comments for Batch Fixing

Group by file — fix all comments in one file at a time

gh ghent comments --pr 42 --group-by file --format json --no-tui

Group by author — address each reviewer's feedback together

gh ghent comments --pr 42 --group-by author --format json --no-tui

Additional Resources

  • Command Reference — all commands, all flags, output schemas

  • Agent Workflows — step-by-step patterns for common tasks

  • Exit Codes — branching logic and conditional patterns

  • Review Cycle Example — full walkthrough: read, fix, resolve, reply

  • CI Monitor Example — watch CI, extract errors, fix, re-check

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Coding

iterm2-driver

No summary provided by upstream source.

Repository SourceNeeds Review
Automation

clinic-visit-prep

帮助患者整理就诊前问题、既往记录、检查清单与时间线,不提供诊断。;use for healthcare, intake, prep workflows;do not use for 给诊断结论, 替代医生意见.

Archived SourceRecently Updated
Automation

changelog-curator

从变更记录、提交摘要或发布说明中整理对外 changelog,并区分用户价值与内部改动。;use for changelog, release-notes, docs workflows;do not use for 捏造未发布功能, 替代正式合规审批.

Archived SourceRecently Updated
Automation

klaviyo

Klaviyo API integration with managed OAuth. Access profiles, lists, segments, campaigns, flows, events, metrics, templates, catalogs, and webhooks. Use this skill when users want to manage email marketing, customer data, or integrate with Klaviyo workflows. For other third party apps, use the api-gateway skill (https://clawhub.ai/byungkyu/api-gateway).

Archived SourceRecently Updated