peer-model

peer-model is a skill for Claude Code from lollipopkit/cc-plugins. It costs 96 tokens per session (1,145 once invoked), scanned A, original, MIT.

A skill for starting another AI model as a separate coding session and exchanging messages with it. The model must use an Anthropic-compatible service, with connection profiles stored in a configuration file.

In plain words
What is it for?
Use it when you need to delegate work to models such as DeepSeek or GLM, optionally share the current conversation, and communicate through peer-session messages.
Why use it?
It lets you ask a different model for help without treating a one-off command-line response as a shared working session.

Skill for Claude Code

Written for Claude Code: allowed-tools in frontmatter. Also seen: mentions Claude Code.

Runs only inside its plugin — its command needs a path that Claude Code sets for a plugin’s own hooks and for nothing else. Install the plugin, not this.

Part of the peer-model plugin — 1 skill shipped together

Good fit Use it when you need to delegate work to models such as DeepSeek or GLM, optionally share the current conversation, and communicate through peer-session messages.

Compare 6 skills from other repositories ↓
Install

Getting it into your agent

This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.

Claude Code
/plugin marketplace add lollipopkit/cc-plugins
Claude Code
/plugin install peer-model

Made for: Claude Code.

Or install peer-model, the plugin that ships this one along with the rest of its 1 skill.

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 peer-model

README.md
[![agentmods](https://agentmods.dev/badge/skills/lollipopkit/cc-plugins/peer-model/github.svg)](https://agentmods.dev/skills/lollipopkit/cc-plugins/peer-model)
Your own site
<a href="https://agentmods.dev/skills/lollipopkit/cc-plugins/peer-model"><img src="https://agentmods.dev/badge/skills/lollipopkit/cc-plugins/peer-model/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 peer-model

Your own site · 80×15
<a href="https://agentmods.dev/skills/lollipopkit/cc-plugins/peer-model"><img src="https://agentmods.dev/badge/skills/lollipopkit/cc-plugins/peer-model.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 96 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,145 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found 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 $0.00096 $0.01145
Opus 5.5 $0.00038 $0.00458
Sonnet 5.5 $0.00019 $0.00229
Haiku 4.5 $0.00010 $0.00114

Measured 7d ago against content hash c3c9ba666a14, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

peer-model 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 7d 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.

plugins/peer-model/skills/peer-model/SKILL.md · 68 lines

How it starts

The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Peer model session

The target model runs as a separate background Claude Code session (claude --bg) pointed at an Anthropic-compatible endpoint. It registers as a local peer, so both sides talk through the built-in SendMessage / ListAgents tools. Do not use claude -p for this: headless print sessions are not peers and cannot message back.

0. Configuration

Profiles live in ${XDG_CONFIG_HOME:-~/.config}/cc-peer-model/config.toml (override with CC_PEER_MODEL_CONFIG). List them with:

"${CLAUDE_PLUGIN_ROOT}/scripts/peer.py" profiles

If the file is missing, run peer.py init to write an example and ask the user to fill it in. Each profile:

[profiles.ds]
base_url_env = "PIMM_BASE_URL"  # or base_url = "...", or base_url_cmd = [...]
api_key_env = "PIMM_API_KEY"    # or api_key_cmd = ["security", "find-generic-password", "-s", "pimm", "-w"]
model = "deepseek-v4.1-flash"
max_context_tokens = 1000000    # optional
env = { API_TIMEOUT_MS = "600000" }  # optional, non-secret only

claude --bg runs the peer in a process spawned by the Claude Code daemon, so the caller's environment never reaches it. peer.py therefore passes the profile via --settings: non-secret values in env, the key via apiKeyHelper, which runs inside the peer session:

  • api_key_cmd: an argv list (e.g. a macOS keychain lookup), run in the session. Works regardless of how the daemon was started.
  • api_key_env: the variable is read in the session, i.e. from the daemon's environment. It works only if the daemon was started from a shell that exports it (e.g. via the shell profile); a daemon started from Claude Desktop, an IDE or launchd will not have it.
  • A literal api_key is refused.

Never write secrets into the config or echo resolved values.

1. Spawn

"${CLAUDE_PLUGIN_ROOT}/scripts/peer.py" spawn <profile> -n <name> [-f] [-m <permission-mode>] [-e <effort>] ["<initial task>"]
  • <profile>: map the user's wording to a profile name (e.g. "deepseek" → ds). If ambiguous, list profiles and ask.
  • -n: peer name, used as the to address. Pick a short task-specific name (e.g. ds-review-auth). Check ListAgents first and reuse an existing peer instead of spawning a duplicate.
  • -f: fork the current session so the peer gets the full conversation so far. Use it whenever the task depends on prior discussion; skip it for self-contained tasks (cheaper, cleaner context). A forked peer also inherits the current task, so its initial prompt must state its scope explicitly and say what not to do.
  • -m: must equal this session's permission mode (default, acceptEdits, plan, bypassPermissions; default bypassPermissions). On a mismatch, messages are held for the user's manual approval instead of being delivered.
  • Initial task: tell it to report back with SendMessage to this session's name (shown by ListAgents as "This session is ..."). Without a task it starts idle and waits for messages.

Read the full file on GitHub · 68 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. 7d ago First seen · 68 lines · 96 tokens per session scan A c3c9ba666a14

Subscribe to this mod's changes

peer-model is a skill published in the GitHub repository lollipopkit/cc-plugins (7 stars, last pushed yesterday), licensed MIT. It adds 96 tokens to every session and 1,145 once invoked, about $0.0004 per session on Opus 5.5. 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 2026-10-07.

Related

Other skills, from other repositories

taiyi-integration

A project workflow skill for closing a TaiyiForge change and recording it in a CHANGELOG.md file. It checks review results, tests, and the state of the working tree before archiving the change.

Dong90/oh-my-taiyiforge · 27 tokens

taiyi-ui-design

A design-planning guide for describing how an application's user interface should look and behave. It produces a UI-DESIGN.md document covering layouts, components, interactions, accessibility, and error states.

Dong90/oh-my-taiyiforge · 35 tokens

taiyi-evolve

A workflow skill that compares the implemented code with the frozen design after development and testing. It records architecture changes and proposes updates to DESIGN.md, the document describing the intended system structure.

Dong90/oh-my-taiyiforge · 37 tokens

taiyi-compress

A workflow tool for shrinking large coding-agent conversations and work files into shorter context notes. It can also coordinate separate agents for parallel development and create handoff notes for continuing work in a new session.

Dong90/oh-my-taiyiforge · 51 tokens

taiyi-diagram-c4

A code-scanning tool that builds C4 architecture documents from a repository. It separates facts observed in the code from conclusions inferred about the design and uses Mermaid diagrams as the source format.

Dong90/oh-my-taiyiforge · 48 tokens

taiyi-ultrawork

A workflow for splitting a coding task into smaller pieces and sending those pieces to parallel coding agents. It also sets file ownership rules so agents do not edit the same files.

Dong90/oh-my-taiyiforge · 39 tokens