Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/g-research/bobbit/client-debugnpx skills add G-Research/bobbit --skill client-debuggit clone --depth 1 https://github.com/G-Research/bobbitWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00037 | $0.01776 |
| Opus 5 | $0.00018 | $0.00888 |
| Sonnet 5 | $0.00007 | $0.00355 |
| Haiku 4.5 | $0.00004 | $0.00178 |
Grade A, and why
client-debug scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 122 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Client debug & mobile/PWA layout playbook
Use this when you need information back from the running web client but can't
get it the usual way — most often an installed mobile PWA (iOS "Add to Home
Screen"), where there is no DevTools console and things like
env(safe-area-inset-*), the iOS viewport "lie", connection state, or which
build is actually live can only be observed on the real device.
The tool: Client debug
A flag-gated diagnostic lives in src/app/client-debug.ts.
- Enable it: Settings → the Debug toggle next to Restart Server in the
footer. It's the unified dev switch (it flips the
bobbit-client-debuglocalStorage flag AND arms boot-timing perf). Dev-harness only — it's gated on the server's harness status (harnessRestartAvailable), so it appears whenever the gateway runs under the dev harness, including on a phone pointed at that dev gateway (which is the whole point); it's hidden on production gateways. - Use it: a small floating DBG button appears in the chat view (desktop and mobile). Tapping it dumps a fenced, copy-pasteable report into the message composer so the user can read it or send it to you.
- Report sections:
Environment(incl.build=id),Viewport / layout,Performance, plus any registered viaregisterDebugSection(name, fn)(e.g.App state, wired fromrender.ts). Add new sections there or via the registry — keepclient-debug.tsgeneric, not bug-specific. - Build id: the DBG button shows a short build id and the report's
Environmentsection printsbuild=…. Use it to confirm the device is running the build you just shipped (the #1 gotcha when iterating on a hard-to- reload installed PWA).
Getting a dump from the user: ask them to enable the flag, tap DBG, and
send the message. On desktop you can instead run collectClientDebug() logic
mentally from browser_eval, but the button is the mobile-friendly path.
Version/identity: the report includes buildId (via the boot waterfall when
present) and live viewport numbers. If you change client code while iterating,
have the user confirm the change took effect by re-dumping — the network-first
service worker means a cold relaunch (while online) pulls the latest; no reinstall
needed.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 122 lines · 37 tokens per session scan A f9da3ac8eddc
client-debug is a skill published in the GitHub repository G-Research/bobbit (11 stars, last pushed 2d ago), licensed MIT. It adds 37 tokens to every session and 1,776 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
orca-emulator-android
Control an Android emulator / device from inside Orca using the orca CLI. Use for listing/booting AVDs, taps, swipes, typing, hardware buttons (incl. Back and Recents), rotation, app install/launch, runtime permissions, the accessibility tree, and logcat — driving a real adb-connected device or emulator.…
speckit-analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
add-sample
Create a SamplesApp sample page with correct theming and attributes. Use when adding UI samples for controls.
mapping-to-snomed
Maps clinical concept spans extracted by OpenMed to SNOMED CT concepts through a USER-SUPPLIED terminology server (the user's own Ontoserver, Snowstorm, or UMLS/UTS), never a bundled vocabulary. Use when the user wants to code findings, disorders, procedures, body structures, or substances to SNOMED CT, run an ECL…
android-tombstone-symbolication
Symbolicate the .NET runtime frames in an Android tombstone file. Extracts BuildIds and PC offsets from the native backtrace, downloads debug symbols from the Microsoft symbol server, and runs llvm-symbolizer to produce function names with source file and line numbers. USE FOR triaging a .NET MAUI or Mono Android app…
auditing-subgroup-fairness
Audit an OpenMed NER or de-identification model for performance disparities across demographic subgroups (sex, age band, race/ethnicity when available) using openmed.eval.fairnessreport. Use when the user wants per-subgroup recall and leakage, wants to check whether de-identification under-protects a group, wants to…