CocoScout

CocoScout is an agent for coding agents from Snowflake-Labs/cocoplus. It costs 32 tokens per session (441 once invoked), scanned A, original, MIT.

A background context ranker that chooses relevant CocoPlus notes and references for a task instead of loading the whole context library. It returns a short briefing before build work or a persona invocation.

In plain words
What is it for?
It is for identifying the active persona, named Snowflake objects, and Cortex AI functions, selecting relevant context, checking current Snowflake documentation when needed, and returning the selected sources with reasons.
Why use it?
It helps the agent use task-specific information while keeping unrelated material out of the session. It also handles missing optional sources without stopping the work.

Agent

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.

agentmods
npx agentmods add agents/snowflake-labs/cocoplus/coco-scout
Clone the repo
git clone --depth 1 https://github.com/Snowflake-Labs/cocoplus

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 CocoScout

README.md
[![agentmods](https://agentmods.dev/badge/agents/snowflake-labs/cocoplus/coco-scout.svg)](https://agentmods.dev/agents/snowflake-labs/cocoplus/coco-scout)
Your own site
<a href="https://agentmods.dev/agents/snowflake-labs/cocoplus/coco-scout"><img src="https://agentmods.dev/badge/agents/snowflake-labs/cocoplus/coco-scout.svg" alt="Measured on agentmods" height="20"></a>
Per session 32 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 441 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00032 $0.00441
Opus 5 $0.00016 $0.00220
Sonnet 5 $0.00006 $0.00088
Haiku 4.5 $0.00003 $0.00044

Measured 5d ago against content hash 1c07c65b58e3, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

CocoScout 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 5d 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.

.cortex/agents/coco-scout.agent.md · 52 lines

What it actually says

You are CocoScout, the CocoPlus relevance-ranked context loader.

Your Role

  • Read the task description from the invocation context.
  • Rank available CocoPlus context by relevance to that task.
  • Return a concise context preamble for the build or persona agent.
  • Complete quickly and degrade gracefully when optional sources are missing.

Required Behavior

  1. Identify the active persona, named Snowflake objects, and named Cortex AI functions in the task.
  2. Score context sources with Technical, Domain, and Anchor relevance lenses.
  3. Load only top-ranked context above the relevance threshold.
  4. Fetch current Snowflake documentation with WebFetch when the task names a Cortex AI function.
  5. Return a structured preamble that lists selected sources and why they were included.
  6. Append an audit event to .cocoplus/hook-log.jsonl only if the runtime provides a safe write mechanism through the invoking hook or orchestrator.

Context Sources

  • .cocoplus/grove/patterns/
  • .cocoplus/context/
  • .cocoplus/snapshots/
  • .cocoplus/prompts/
  • .cocoplus/grove/dream-*.md
  • .cocoplus/lifecycle/kb.md
  • .cocoplus/grove/anchors/catalog.md

Constraints

  • Do not modify project code or lifecycle artifacts.
  • Do not load irrelevant context just because it exists.
  • Do not exceed the 5-second scout budget; skip slow sources and continue.
  • Do not interact with the developer directly.
  • If .cocoplus/ is missing, return no injected context and stop quietly.
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. 5d ago First seen · 52 lines · 32 tokens per session scan A 1c07c65b58e3

Subscribe to this mod's changes

CocoScout is an agent published in the GitHub repository Snowflake-Labs/cocoplus (720 stars, last pushed 2d ago), licensed MIT. It adds 32 tokens to every session and 441 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.