phone-a-team

phone-a-team is a skill for Claude Code, Codex from freibergergarcia/phone-a-friend. It costs 75 tokens per session (2,465 once invoked), scanned C, a copy of phone-a-team, Apache-2.0.

A workflow that asks several other AI systems to review a task across multiple rounds, combining their replies between rounds. It supports Claude, Gemini, OpenCode, and Ollama backends.

In plain words
What is it for?
Use it when a task benefits from multi-model critique and iterative refinement rather than a single outside opinion.
Why use it?
Repeated independent feedback can expose weaknesses and refine an answer before it is finalized.

Skill for Claude CodeCodex

Written for Claude Code and Codex: argument-hint in frontmatter, but also installed under .codex/. Also seen: mentions subagents; positional $N argument; mentions Codex.

Part of the phone-a-friend plugin — 4 skills, 3 commands shipped together

Good fit Use it when a task benefits from multi-model critique and iterative refinement rather than a single outside opinion.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/freibergergarcia/phone-a-friend/codex
Install

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.

Any agent
npx skills add freibergergarcia/phone-a-friend --skill codex
Clone the repo
git clone --depth 1 https://github.com/freibergergarcia/phone-a-friend

Made for: Claude Code, Codex.

Or install phone-a-friend, the plugin that ships this one along with the rest of its 4 skills, 3 commands.

Wrote 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.

agentmods badge for phone-a-team

README.md
[![agentmods](https://agentmods.dev/badge/skills/freibergergarcia/phone-a-friend/codex/github.svg)](https://agentmods.dev/skills/freibergergarcia/phone-a-friend/codex)
Your own site
<a href="https://agentmods.dev/skills/freibergergarcia/phone-a-friend/codex"><img src="https://agentmods.dev/badge/skills/freibergergarcia/phone-a-friend/codex/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.

agentmods 80×15 button for phone-a-team

Your own site · 80×15
<a href="https://agentmods.dev/skills/freibergergarcia/phone-a-friend/codex"><img src="https://agentmods.dev/badge/skills/freibergergarcia/phone-a-friend/codex.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 75 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,465 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 1 finding. A grade says what 26 rules found in the file — not that it is safe.
Origin 100% copy Near-identical to another mod in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00075 $0.02465
Opus 5 $0.00037 $0.01233
Sonnet 5 $0.00015 $0.00493
Haiku 4.5 $0.00007 $0.00247

Measured 9d ago against content hash dd57da2d5f14, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade C, and why

phone-a-team scanned grade C with 1 finding 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 9d 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.

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 "$ROUND_DIR"' EXIT
Origin

This is a copy

100% identical to phone-a-team — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.

skills/phone-a-team/.codex/SKILL.md · 180 lines

How it starts

The opening of the file, as written. The whole thing — 180 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/phone-a-team for Codex

Iterative multi-model refinement loop. Use this when the user wants more than a single relay — they want backends to push back on each other across rounds and converge on something. For one-shot "ask X and Y" requests, use /phone-a-friend (which already supports parallel multi-backend) instead.

You orchestrate the rounds yourself via Bash. Do not spawn Codex subagents for this — the skill is designed to run end-to-end in the parent turn. Codex's own model does the synthesis between rounds by reading the per-backend outputs you write to disk.

Inputs

$ARGUMENTS contains the user's request. Parse:

  • TASK_DESCRIPTION: free-form text with flags stripped.
  • --backend <name> (or --backend a,b): comma-separated list of friend backends to query each round. Allowed: claude, gemini, opencode, ollama. Never codex (the recursion guard refuses; you would be calling yourself). Default: claude,gemini. antigravity is deliberately excluded until the PaF backend supports session continuity; /phone-a-friend supports one-shot Antigravity relays.
  • --max-rounds N: rounds before giving up. Clamp to [1, 5]. Default: 3.
  • --model <name>: only meaningful for ollama and opencode.

If the user named only one backend, the team degenerates to a self-review loop against that backend across rounds. That is valid; do not error out, just proceed.

Step 0 — Preflight

  1. Resolve the binary. Always use the PATH-installed phone-a-friend, never ./phone-a-friend from the current directory.

    RELAY_BIN="$(command -v phone-a-friend)"
    if [ -z "$RELAY_BIN" ]; then
      echo "phone-a-friend binary not on PATH. Install with: npm i -g @freibergergarcia/phone-a-friend" >&2
      exit 1
    fi
    
  2. Resolve the repository path.

    REPO="$(git rev-parse --show-toplevel 2>/dev/null || pwd)"
    
  3. Check the Codex sandbox mode. You are running inside Codex. By default, Codex uses workspace-write, which blocks subprocess access to the macOS Keychain and most outbound network. That means relays to Claude will return Not logged in · Please run /login (false alarm — Claude is logged in; the sandbox is intercepting) and relays to Gemini will hang until the timeout fires (Gemini cannot reach Google for OAuth refresh).

Read the full file on GitHub · 180 lines

Changes

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.

  1. 9d ago First seen · 180 lines · 75 tokens per session scan C dd57da2d5f14

Subscribe to this mod's changes

phone-a-team is a skill published in the GitHub repository freibergergarcia/phone-a-friend (10 stars, last pushed yesterday), licensed Apache-2.0. It adds 75 tokens to every session and 2,465 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it C with 1 finding (recursive force delete). It is 100% identical to phone-a-team, differing in 0 lines, and is treated as a copy.

Related

Other skills, from other repositories

roundtable

Convene a "round table" of multiple frontier AI models — Grok, Codex/OpenAI, GLM, MiniMax, Claude, and Gemini — to get independent perspectives on a question, optionally deliberating across rounds until a Claude chair declares consensus, then synthesize a single recommendation. Use when the user says "round table"…

frontier-infra/roundtable · 168 tokens

save-learning

Saves user instructions as persistent learnings for future sessions. Use when the user says 'remember this', 'always do X', 'from now on', 'never do Y', or gives any instruction they want persisted across sessions. Proactively suggest when the user states a preference, convention, or rule they clearly want followed in…

caliber-ai-org/ai-setup · 71 tokens

gwm

Manage git worktrees across any repository with the gwm Rust binary (CLI + ratatui TUI). Use when the user asks to create / list / remove / bootstrap / switch / link worktrees, run a command across worktrees (gwm exec) or reclaim build artifacts (gwm clean), materialise a PR into a worktree (gwm review), drive a…

kbrdn1/gwm-cli · 0 tokens

web-access

A method for handling difficult web tasks across search, page fetching, and a real browser. It covers sites that need login sessions, dynamic pages, anti-bot handling, or exploration before their structure is understood.

open-octo/octo-agent · 139 tokens

skill-creator

A skill for creating or improving reusable instructions for an agent. It covers capturing the intended workflow, writing a draft, testing it, and refining its trigger description.

open-octo/octo-agent · 123 tokens

packing-list

A travel-packing aid that builds a checklist from the destination's climate, trip type, length, transport, accommodation, and travelling group. It treats a city break, hike, business trip, and family holiday as different packing problems.

open-octo/octo-agent · 105 tokens