Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/dorukardahan/headless-relaynpx agentmods add skills/dorukardahan/headless-relay/headless-relayWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/dorukardahan/headless-relay/headless-relay)<a href="https://agentmods.dev/skills/dorukardahan/headless-relay/headless-relay"><img src="https://agentmods.dev/badge/skills/dorukardahan/headless-relay/headless-relay/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/dorukardahan/headless-relay/headless-relay"><img src="https://agentmods.dev/badge/skills/dorukardahan/headless-relay/headless-relay.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What 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 5.1 | $0.00165 | $0.21458 |
| Opus 5 | $0.00082 | $0.10729 |
| Sonnet 5 | $0.00033 | $0.04292 |
| Haiku 4.5 | $0.00016 | $0.02146 |
Grade D, and why
headless-relay scanned grade D with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
`~/.codex/config.toml` (`sandbox_mode`, `approval_policy`, reviewer features), so headless Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
trap 'rm -rf "$base" "${sbx:-}" 2>/dev/null' EXIT # armed BEFORE sandbox exists: nothing can leak How it starts
The opening of the file, as written. The whole thing — 890 lines — stays where its author put it; the contents beside it link to each section on GitHub.
headless-relay
This skill provides instructions for delegating a task to another AI model without leaving the current agent session. The user says "ask Codex", "get GLM's opinion", "run this by Grok"; the orchestrating agent (Claude Code, Codex CLI, or another harness) writes the prompt, runs the target model's headless CLI through its shell tool, reads the model's stdout, and summarizes the answer back. Follow these patterns exactly.
The five targets
| Target | CLI | Headless entry point | Auth / plan |
|---|---|---|---|
| GPT | codex (OpenAI Codex CLI) |
codex exec |
ChatGPT plan or OpenAI API key (codex login) |
| GLM | opencode, or zcode (ships inside the ZCode desktop app) |
opencode run / zcode --prompt |
Z.ai Coding Plan (API key or ZCode app login) |
| Grok | grok (xAI Grok Build) |
grok -p — via the grok_relay helper (empty HOME + clean temp GROK_HOME so it can't read your other tools' config or grok's own global rules; see the next section) |
SuperGrok (grok login) or XAI_API_KEY |
| Gemini | agy (Google Antigravity CLI — replaced the retired Gemini CLI) |
agy -p / --print |
Google account via the Antigravity app/CLI |
| Claude | claude (Claude Code) |
claude -p, or the harness's native subagent |
Anthropic auth of the current session |
Grok: read this before relaying
Grok Build had a real data-egress problem, and it has since changed. Both halves matter.
What happened. Earlier shipped versions of grok, run inside a git repository, uploaded your
entire tracked repo — full commit history and every tracked file, including a tracked .env —
to xAI cloud storage as a git bundle, independent of which files the model read, not stopped by any
prompt or flag. Confirmed by xAI's own Grok account on X and by independent mitmproxy wire capture
(a never-read canary file was recovered from the upload). Details: SECURITY.md.
What changed (2026-07-15). xAI open-sourced Grok Build (Apache-2.0,
github.com/xai-org/grok-build), deleted previously-retained coding data, and set retention off by
default. A source audit of that release (commit c68e39f) found:
- The whole-repo bundle path is gone from the source — no
git bundle/codebase_upload/ whole-repo archive anywhere; the trace pipeline serializes only the model's own turn I/O, and it defaults off (telemetry defaults off; the per-turnconfig_files.jsonupload is hard-disabled). - Your local
~/.grok/config.tomlbeats xAI's remote settings (proven by the resolver plus a repo test): a server-pushed flag cannot re-enable an upload you turned off locally.
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.
- 10d ago First seen · 890 lines · 165 tokens per session scan D 93a7044587dd
headless-relay is a skill published in the GitHub repository dorukardahan/headless-relay (2 stars, last pushed 26d ago), licensed MIT. It adds 165 tokens to every session and 21,458 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it D with 2 findings (reads agent configuration directories, recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
claude-md-improver
Audit and improve CLAUDE.md files in repositories. Use when user asks to check, audit, update, improve, or fix CLAUDE.md files. Scans for all CLAUDE.md files, evaluates quality against templates, outputs quality report, then makes targeted updates. Also use when the user mentions "CLAUDE.md maintenance" or "project…
agent-platform-rag-engine-management
Manage and query Agent Platform RAG Engine Corpora and retrieve grounded contexts using the Google GenAI SDK. Use when listing RAG corpora or files, inspecting a corpus, retrieving contexts, or generating content grounded in a RAG corpus. Do not use for standard database queries (use SQL/Spanner skills), Google…
gke-workload-security
Audits, configures, and hardens workload-level security controls for Google Kubernetes Engine (GKE) applications and namespaces. Covers running cluster security audits (auditcluster.sh), configuring Workload Identity Federation (impersonation, KSA/GSA binding, and pod setup), enforcing Network Policies (default-deny…
gke-reliability
Improves GKE workload reliability, using PDBs, health probes, and topology spread constraints. Use when configuring GKE workload reliability, setting up PDBs, or configuring GKE health probes (liveness, readiness, startup). Don't use for disaster recovery setup or full cluster backups (use gke-backup-dr instead).
agent-platform-model-registry
Agent Platform Model Registry Management. Use when you need to upload, list, describe, update, or delete machine learning models (and their versions) in the Agent Platform Model Registry. Don't use for model training, model deployment to endpoints, or managing non-Agent Platform models.
google-cloud-solution-agentic-analytics-spark-knowledge-catalog
Discovers requirements and generates guidance to design and deploy a governed, secure agentic-analytics solution for data that's distributed across Google Cloud, other cloud providers, or on-premises. Data that's outside Google Cloud (such as data from Databricks, Snowflake, Salesforce, SAP, or Oracle systems) is…