bonedigger 🦴

ujust report filing + confirm-driven priority escalation, using GitHub as the message bus.

Current scope

Lifecycle management moved to projectbluefin/common/.github/workflows/lifecycle.yml.

bonedigger now handles only:

• ujust report issue detection on open
• confirm-count priority escalation (3+ → priority/p1, 5+ → priority/p0)
• bonedigger-specific agent donation fast-track labels on issue open

Owned by common now:

• slash commands like /approve, /claim, /unclaim, /wontfix, /hold
• issue body widget rendering
• label creation / sync
• stale sweeps and lifecycle transitions

How it works

USER'S MACHINE                    GITHUB
─────────────────                 ─────────────────────────────────────────
ujust report                      bonedigger lifecycle workflow
  └─ collects diagnostics           └─ detect `ujust report` issue bodies
  └─ PII scrub on-device            └─ keep confirm-based priority labels in sync
  └─ user reviews locally           └─ fast-track agent donation issues
  └─ uploads to user's gist
  └─ opens issue w/ gist link     common lifecycle workflow
                                   └─ slash commands + widget + queue state
ujust confirm <issue#>            └─ bonedigger re-counts confirms
                                     and escalates priority labels

GitHub Issues is the only backend. No central server. User owns their data.

Usage

As a user

Run on your Bluefin machine:

ujust report       # file a bug report
ujust confirm 42   # confirm you hit issue #42 too
ujust verify 42    # verify issue #42 is fixed after an update

As a repo maintainer

Add to .github/workflows/bonedigger.yml:

name: bonedigger
on:
  issues:
    types: [opened]
  issue_comment:
    types: [created]

permissions:
  issues: write
  contents: read

jobs:
  bonedigger:
    uses: projectbluefin/bonedigger/.github/workflows/lifecycle.yml@main
    secrets: inherit

If you also want lifecycle widgets, slash commands, and queue management, wire the caller from projectbluefin/common/.github/workflows/lifecycle.yml too.

Repository structure

• just/ — canonical ujust report recipe (shipped via projectbluefin/common)
• templates/ — canonical GitHub issue templates (shipped to all org repos)
• .github/workflows/lifecycle.yml — reusable reporting workflow
• action.yml — composite action entrypoint

Privacy

• All PII scrubbing happens on the user's machine before any upload
• Diagnostic gists belong to the user — bonedigger only reads them, never creates its own
• No central server, no telemetry infrastructure required

Roadmap

Planned: crash/panic detection in ujust report

The diagnostic collector currently captures a live system snapshot but has no awareness of what happened in the previous boot. A full class of bugs — kernel panics during sleep, hard lockups, abrupt reboots — leave zero trace in the current session.

Planned work:

• #11 — crash/panic detection section: unclean boot classifier (4 buckets: clean shutdown / suspend-no-resume / abrupt end / journal unavailable), panic keyword scan of previous boot, crash artifact status (pstore, kdump, coredumps)
• #12 — PII scrubbing for kernel log excerpts: IPv4/IPv6, UUIDs, disk serials, MAC addresses

Part of Project Bluefin

• projectbluefin/common — ships ujust report and owns lifecycle management
• projectbluefin/dakota — reference implementation