All skills
block avatar

/berd-help

@9b88cac
by Block Open Sourceblock/berd955 stars
124

Help with Berd the desktop app: how-to, troubleshooting, settings, agents, skills, automations, projects, sessions, providers, connections, feedback, or berdctl. Use for app questions, not harness-specific behavior.

  • 11 files
  • 19.2 KB
  • Updated 2 months ago
  • GitHub

Use this Skill: https://skilld.dev/gh/block/berd/berd-help

This session only. Nothing lands on disk.

referencesreporting-problems.md

≈641 tokens on demand. Your agent reads this file only when SKILL.md points to it.

Reporting Problems

Route app-level bugs, crashes, and feedback through an available feedback route rather than improvising a report. Routes vary by build — verify which exist before recommending one, and use the first that applies:

  1. Run berdctl feedback --help. If the feedback subcommand exists, follow the guidance below. If invoking it reports that feedback is unavailable or disabled, continue to the next route.
  2. Otherwise, check whether the app shows a visible in-app feedback action (labeled "Send feedback"). If it's visible, point the user there — don't assume it's present.
  3. Otherwise, the fallback is the public issue tracker: https://github.com/block/berd/issues. You may help the user draft a report (title, description, steps to reproduce, app version, OS), but the user reviews and files it themselves. Never file on their behalf, and never invent another URL or email for reports.

When berdctl feedback exists:

  • Prefer berdctl feedback open --title <title> --description <description> (add --include-logs only when the user wants diagnostics attached). This opens the same feedback form the UI's "Submit feedback" action opens, with the report prefilled — nothing is sent. The user reviews, edits, adds image attachments, and submits it themselves, exactly as if they had filled out the dialog by hand.
  • berdctl feedback submit files the report immediately, with no review step and no confirmation screen — the moment it runs, a real ticket exists. Only use it when the user has explicitly and unambiguously asked to file, send, or submit feedback right now, matching the command's own guidance. Default to feedback open whenever there's any ambiguity about whether the user wants to review first.
  • State the privacy posture accurately when asked: attaching logs (via --include-logs, either command) includes safe local metadata and sanitized logs. It excludes prompts, responses, tool payloads, and session databases — those are never attached. Logs are opt-in only; omitting --include-logs attaches nothing.
  • Do not invent your own log bundle, file path, or diagnostic procedure. If a user needs something the feedback flow does not cover, say what it does cover and stop there rather than proposing a workaround that bypasses redaction.
  • If the issue is about a specific harness's behavior (not the Berd app), say that it's out of scope for this flow and point them to the harness's own reporting channel if you know it — don't guess.

Source: SKILL.md on GitHub

No third-party reports yet.

Signed by skilld at 9b88cac. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 3 days ago.

Activeupdated 2 months ago
metadata
{
  "berdBundled": true
}

README badge

README badge for block/berd/berd-help