replen-onboard

replen-onboard is a skill for Claude Code, Codex from replenhq/replen. It costs 38 tokens per session (6,834 once invoked), scanned A, original, Apache-2.0.

A one-time setup skill that studies the repositories you choose and builds profiles of their capabilities. It can also quietly update the profile for one repository when needed.

In plain words
What is it for?
Use it when setting up Replen across several projects or when a single new or changed repository needs its capabilities recorded.
Why use it?
Generic matching can suggest irrelevant results when the agent does not understand a project's code and tools. This setup grounds later recommendations in the repositories you use.

Skill for Claude CodeCodex

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 skills/replenhq/replen/replen-onboard
Any agent
npx skills add replenhq/replen --skill replen-onboard
Clone the repo
git clone --depth 1 https://github.com/replenhq/replen

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 replen-onboard

README.md
[![agentmods](https://agentmods.dev/badge/skills/replenhq/replen/replen-onboard.svg)](https://agentmods.dev/skills/replenhq/replen/replen-onboard)
Your own site
<a href="https://agentmods.dev/skills/replenhq/replen/replen-onboard"><img src="https://agentmods.dev/badge/skills/replenhq/replen/replen-onboard.svg" alt="Measured on agentmods" height="20"></a>
Per session 38 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 6,834 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. 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 $0.00038 $0.06834
Opus 5 $0.00019 $0.03417
Sonnet 5 $0.00008 $0.01367
Haiku 4.5 $0.00004 $0.00683

Measured 3d ago against content hash 5a11c1197fcd, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

replen-onboard scanned grade A 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 3d 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

are judged against the mission. Use the MCP tools, not hand-rolled `curl`.
cli/extras/skills/replen-onboard/SKILL.md · 451 lines

How it starts

The opening of the file, as written. The whole thing — 451 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Replen onboarding — autonomous, background, multi-repo setup

You are setting Replen up properly across the user's projects. The better the grounding you build now, the sharper Replen's matches are forever after — and it all happens in-session, on the user's own agent, so nothing leaves their machine except the project profile you build.

This skill is designed to run AUTONOMOUSLY. After ONE upfront confirmation of the repo list, do NOT stop to ask questions — every policy decision is fixed below. The user should be able to kick this off and keep working (or let a background agent run it) without babysitting it.

Two modes: full onboard vs silent delta re-ground

  • Full onboard (default — the user typed /replen-onboard): profile all the in-scope repos as laid out below, starting with the Step 0 brief.
  • Silent first-time ground of ONE repo (you were sent here by the background session-start flow to profile a single unprofiled repo): do NOT show the Step 0 brief, do NOT run Step 1 discovery, do NOT touch other repos. Profile just THIS repo silently — jump to reading its code (Step 2 onward) and push replen_set_capabilities (mode: "replace", its first full set) + replen_set_versions + replen_set_tags. Stay silent throughout.
  • Silent delta re-ground (triggered when replen_match / replen_onboard_state returns needsReground: true for a SINGLE, already-grounded repo — the code drifted or the grounding schema advanced): this is a scoped refresh, NOT a full re-onboard. Do NOT show the Step 0 brief, do NOT ask, do NOT touch other repos, do NOT re-read the whole codebase. Run it in a BACKGROUND subagent and stay silent. Steps:
    1. git diff --name-only <grounded-sha>..HEAD — the grounded SHA is on the repo's grounding fingerprint; if it's unavailable, inspect what changed recently.
    2. Read only the changed / new areas.
    3. Re-derive just the NEW or CHANGED capabilities — new domains, new mechanism (how it's built), updated maturity (a capability that gained a real library dependency flips hand-rolledlibrary-backed; a new from-scratch module is a new hand-rolled capability).
    4. Call replen_set_capabilities with mode: "merge" (never replace) for just those capabilities. That re-stamps the grounding fingerprint and clears the signal. Add file paths for anything new. Refresh replen_set_versions if the lockfile moved.

Read the full file on GitHub · 451 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. 3d ago First seen · 451 lines · 38 tokens per session scan A 5a11c1197fcd

Subscribe to this mod's changes

replen-onboard is a skill published in the GitHub repository replenhq/replen (8 stars, last pushed 24d ago), licensed Apache-2.0. It adds 38 tokens to every session and 6,834 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). 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

foolproof-product

Use this skill on ANY code change in codevira's product surface (mcpserver/, indexer/). Triggers on Edit/Write to product code, on phrases like "fix bug", "add feature", "handle error", or any modification to commands the end-user runs. Enforces 10 product- reliability principles: no silent failures, atomic state…

sachinshelke/codevira · 124 tokens

epistemic-honesty

Use this skill whenever you're about to (1) make a definitive claim like "this works", "tests pass", "ready to ship", "fixed", "done"; or (2) propose a solution to a problem; or (3) diagnose a bug or issue. Forces explicit confidence calibration with inline evidence, surfaces what you don't know, and demands a…

sachinshelke/codevira · 106 tokens

open-source-quality

Use this skill before committing, pushing, or opening a PR. Triggers on phrases like "commit", "git commit", "push", "open PR", "create pull request", "finalize this change". Enforces conventional commit messages, atomic commits, code style (ruff/mypy), public-API docstrings, actionable error messages, changelog…

sachinshelke/codevira · 101 tokens

release-readiness

Use this skill whenever the conversation mentions releasing, shipping, publishing, promoting to PyPI, "ship it", "release X.Y.Z", "ready to release", "let's publish", or any phrase suggesting moving codevira to production. Walks the 5-gate release gauntlet (G1-G5) and refuses to proceed without evidence at every gate.…

sachinshelke/codevira · 96 tokens

development-discipline

Use this skill before ANY code-writing tool call (Edit, Write, NotebookEdit) on this project. Triggers on phrases like "implement", "fix bug", "add feature", "refactor", "create function", "modify file", or whenever the user asks for a code change. Forces a 4-step CONTEXT → PURPOSE → REASON → CODE sequence with…

sachinshelke/codevira · 129 tokens

django-patterns

Django architecture patterns, REST API design with DRF, ORM best practices, caching, signals, middleware, and production-grade Django apps.

deepelementlab/jupyter-studio · 32 tokens