Skip to content
Agentmart
Submit a tool

conscientious-objector

Use when an IDE agent (Cursor, Codex, Claude Code) is operating inside BACKS / BACKSAIOS and a human-quality watchdog is needed to verify the harness is being…

Skill by Tcuzzo

IndexedMIT★ 2
Write a review

Chalk Labs

Need help deploying conscientious-objector?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/tcuzzo-conscientious-objector

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Use when an IDE agent (Cursor, Codex, Claude Code) is operating inside BACKS / BACKSAIOS and a human-quality watchdog is needed to verify the harness is being used truthfully. The Conscientious Objector flips the build RED and strips tools from any agent that bypasses the harness (missing skill invocation, mock-theater tests, claim-vs-code drift, silent fallbacks, hidden restrictions). It NEVER sides with an agent going outside the runtime. Sentry-locked: any shape change is RED-FLAGged; the operator can disarm via dual-factor (authenticator TOTP + Telegram approval button) within a 90-second window. Trigger words: CO, conscientious objector, harness watchdog, harness truthfulness, harness bypass, going outside the runtime, harness veto, RED flip, tool strip, RED FLAG, dual-factor kill switch, sentry lock, BACKSAIOS IDE watchdog.

Use cases: agent-building

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/conscientious-objector

What's inside

Skill name
conscientious-objector
Description
Use when an IDE agent (Cursor, Codex, Claude Code) is operating inside BACKS / BACKSAIOS and a human-quality watchdog is needed to verify the harness is being used truthfully. The Conscientious Objector flips the build RED and strips tools from any agent that bypasses the harness (missing skill invocation, mock-theater tests, claim-vs-code drift, silent fallbacks, hidden restrictions). It NEVER sides with an agent going outside the runtime. Sentry-locked: any shape change is RED-FLAGged; the operator can disarm via dual-factor (authenticator TOTP + Telegram approval button) within a 90-second window. Trigger words: CO, conscientious objector, harness watchdog, harness truthfulness, harness bypass, going outside the runtime, harness veto, RED flip, tool strip, RED FLAG, dual-factor kill switch, sentry lock, BACKSAIOS IDE watchdog.
License
MIT
Instructions
45 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Conscientious Objector — the BACKS harness watchdog

This is the AIOS-pack load-side surface. The full implementation lives in the
BACKS runtime at `skills/conscientious-objector/SKILL.md` and
`backend/services/conscientious_objector.py` — load it from there when this
session is running inside the BACKS repo.

## When the CO is the right organ

Any agent building with BACKSAIOS in an IDE — Cursor, Codex, or Claude Code —
needs a watchdog that:

- Refuses to side with the agent when the harness is bypassed.
- Detects harness bypass in real time (real probes, not hardcoded rules).
- Flips the build RED and strips tools from the watched agent on a real finding.
- Locks its own shape against silent change (sentry).

That watchdog is the Conscientious Objector. Attach it; do not bypass it.

## The four kill switches the operator can fire (on-site, no Telegram required)

1. **Surface file flip** — write `{"override": true}` to
   `backend/state/co/operator_override.json`.
2. **IDE / chat SHA code** — type `CO-OVERRIDE: <64-hex-sha>` in the chat.
3. **Local env token** — set `CO_OPERATOR_LOCAL_TOKEN` in `.env` to the
   operator-issued token (60-second window).
4. **Telegram dual-factor** — operator authenticator TOTP plus the Approve
   button on the CO's Telegram card.

All four must work. The CO is forbidden to lock out the operator.

## What this pack skill is, and what it isn't

- **Is** — the AIOS-portable load-side surface. Lets the IDE discover the CO
  via this pack and route the operator's `/co` invocation to the BACKS
  implementation.
- **Isn't** — the watchdog itself. The watchdog is the BACKS-side
  `ConscientiousObjector` class. This skill is the address; that class is the
  body.

When this session is inside the BACKS repo, the runtime resolves the CO from
the BACKS source. When this session is outside BACKS, this skill is the
trailer that says "use BACKS to get the CO."

From Tcuzzo/HydraAgent_public, MIT.

More from Tcuzzo/HydraAgent_public

Signals

Stars
2
Forks
0
Open issues
0
Last commit (this folder)
Oct 6, 2026 (25 days ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from Tcuzzo

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get conscientious-objector running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying conscientious-objector?

CallGet help
Cite this page

conscientious-objector skill. Agent Mart. https://agentmart.dev/skills/tcuzzo-conscientious-objector (last updated 2026-10-24)