seed-curator

seed-curator is a skill for Claude Code, Codex from DojoGenesis/mcp. It costs 31 tokens per session (756 once invoked), scanned A, original, MIT.

A skill for managing reusable development insights, called seeds, across their lifecycle. It can plant, harvest, search, elevate, and audit these insights.

In plain words
What is it for?
Use it to capture a pattern from an experience, scan a session for useful lessons, find relevant seeds, promote proven seeds, or check the seed library.
Why use it?
It gives recurring lessons from sessions or projects a structured place to be recorded, found, reviewed, and promoted into reusable skills.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: model in frontmatter.

Good fit Use it to capture a pattern from an experience, scan a session for useful lessons, find relevant seeds, promote proven seeds, or check the seed library.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/dojogenesis/mcp/seed-curator
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 DojoGenesis/mcp --skill seed-curator
Clone the repo
git clone --depth 1 https://github.com/DojoGenesis/mcp

Made for: Claude Code, Codex.

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 seed-curator

README.md
[![agentmods](https://agentmods.dev/badge/skills/dojogenesis/mcp/seed-curator.svg)](https://agentmods.dev/skills/dojogenesis/mcp/seed-curator)
Your own site
<a href="https://agentmods.dev/skills/dojogenesis/mcp/seed-curator"><img src="https://agentmods.dev/badge/skills/dojogenesis/mcp/seed-curator.svg" alt="Measured on agentmods" height="20"></a>
Per session 31 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 756 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 $0.00031 $0.00756
Opus 5 $0.00015 $0.00378
Sonnet 5 $0.00006 $0.00151
Haiku 4.5 $0.00003 $0.00076

Measured 7d ago against content hash 6aefa9349662, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

seed-curator 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.

internal/skills/bundled/dojo-craft/seed-curator/SKILL.md · 81 lines

How it starts

The opening of the file, as written. The whole thing — 81 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Seed Curator

Manages the full seed lifecycle — from planting raw patterns through harvesting session insights to elevating proven seeds into reusable skills.

Workflow

1. MODE

Determine the operation from user intent:

  • plant — Extract a new seed from a specific experience or insight
  • harvest — Scan a recent session for candidate seeds
  • search — Find seeds relevant to a current situation
  • elevate — Promote a proven seed to a full SKILL.md
  • audit — Check seed library health

2. PLANT (when mode = plant)

Extract a reusable pattern:

  • Name: Short, memorable, hyphenated (e.g., compilation-as-contract)
  • Source: Where this pattern was observed (session, project, incident)
  • Confidence: 0.0-1.0 based on how many times it's been validated
  • Tags: 3-5 categorization tags
  • Pattern: What the seed captures — the reusable insight
  • When to use: Situations where this seed applies
  • Example: Concrete instance of the pattern in action
  • Counter-example: When this pattern does NOT apply Write as YAML-fronted markdown file.

3. HARVEST (when mode = harvest)

Scan the current or recent session for seed candidates:

  • Corrections received from the user (feedback → pattern)
  • Approaches that worked unexpectedly well (success → pattern)
  • Non-obvious decisions that required investigation (discovery → pattern)
  • Friction points that were resolved (obstacle → pattern) Present candidates to user for confirmation before writing files.

4. SEARCH (when mode = search)

Given a situation description, score all seeds by relevance:

  • Tag match (exact tag = 3 points, partial = 1)
  • Description keyword overlap (each matching word = 1 point)
  • Confidence weighting (score × confidence) Return ranked list with seed name, confidence, and one-line relevance explanation.

5. ELEVATE (when mode = elevate)

For seeds meeting promotion criteria:

  • Confidence ≥ 0.8
  • Applied 3+ times across different contexts
  • No contradicting seeds in the library Generate a SKILL.md draft following the skill-creation skill's format:
  • YAML frontmatter with name, description, triggers, version
  • Structured workflow steps derived from the seed's pattern
  • Anti-patterns derived from the seed's counter-examples

Read the full file on GitHub · 81 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 · 81 lines · 31 tokens per session scan A 6aefa9349662

Subscribe to this mod's changes

seed-curator is a skill published in the GitHub repository DojoGenesis/mcp (0 stars, last pushed 1mo ago), licensed MIT. It adds 31 tokens to every session and 756 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 2026-08-31.

Related

Other skills, from other repositories