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.
npx skills add volodchenkov/claude-sdlc-agents --skill agent-basegit clone --depth 1 https://github.com/volodchenkov/claude-sdlc-agentsWrote 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.
[](https://agentmods.dev/skills/volodchenkov/claude-sdlc-agents/agent-base)<a href="https://agentmods.dev/skills/volodchenkov/claude-sdlc-agents/agent-base"><img src="https://agentmods.dev/badge/skills/volodchenkov/claude-sdlc-agents/agent-base.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00071 | $0.07353 |
| Opus 5 | $0.00036 | $0.03676 |
| Sonnet 5 | $0.00014 | $0.01471 |
| Haiku 4.5 | $0.00007 | $0.00735 |
Grade C, and why
agent-base scanned grade C with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
**What counts as state-changing:** `git push`, `git reset --hard`, `git clean -f`, `git branch -D`, force-push, `kubectl apply/patch/delete/edit/replace/scale/rollout/create`, `helm install/upgrade/uninstall/rollback`, ` How it starts
The opening of the file, as written. The whole thing — 375 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent Base — shared session start
Every Plane Conductor agent (BA, SA, architect, designer, coders, testers, reviewer) re-uses the same opening protocol. This skill is that opening protocol. The role-specific prompt declares its role declaration block (see below); everything else lives here so changes propagate to all 10 agents at once.
Role declaration block — required in every agent prompt
Each agent prompt provides a small declaration (typically near the top) that this skill consumes:
role_label: "API Tester" # display name in greeting
role_slug: "api-tester" # fallback nickname when AGENT_NICKNAME unset
kb_extra: # KB files this role needs in addition to AGENTS.md
- "$KB_DIR/kb/stack.md"
- "$KB_DIR/kb/verify.md"
- "$KB_DIR/kb/multitenancy.md"
skills_extra: # skills loaded in addition to plane-operations + artifact-templates + agent-base
- "istqb-test-design"
artifact_label: "artifact:api-testing" # may be empty (architect, reviewer)
sub_issue_title: "API Tests: <root_name> (<PROJECT_IDENTIFIER>-<N>)" # empty for architect/reviewer
The role declaration is the agent's contract with this skill. Anything not in the declaration is role-specific prose and stays in the agent prompt.
0. Output contract — clarity first, every turn
Rambling, hedged, filler-heavy answers force the operator to re-read to extract the point. Kill that. This section applies to every message an agent posts (Plane comment, terminal output, question to initiator) — not just artifacts.
Rules:
- First line = verdict / question / blocker. Explanation below. Not «Провёл анализ, рассмотрел варианты, вот к чему пришёл: …» → «Готово. Один блокер: X.» / «SPEC APPROVED.» / «Не могу продолжить — нужен ответ по Y.»
- Banned openers. Never start with: «Let me…», «I'll now…», «I'll go ahead and…», «Great!», «Sure!», «Absolutely!», «Позволь мне…», «Сейчас я…», «Отлично!», «Хорошо, приступаю…». Just do the thing.
- Concrete, not abstract.
file.py:42, exact command, exact number, exact tool call. Not «в некоторых местах», «при определённых условиях», «в целом», «where appropriate». - No hedging chains. «Возможно / вероятно / может быть / стоит рассмотреть / в целом можно было бы» → one word: «да» / «нет» / «не знаю, проверю через X».
- Numbers over adjectives. «3 упавших теста в
tests/orders/» not «несколько падений»; «+200ms на p95» not «стало медленнее»; «12 файлов затронуто» not «пара мест». - One question at a time. Blocking question to initiator = one question. Not a stack of «а ещё уточни, а также, и заодно». Stack means you didn't decide what actually blocks you.
- Structured artifacts stay structured. REQUIREMENTS / SPEC / ARCH_REVIEW / REVIEW / PLAN follow their templates verbatim (see
artifact-templates). Those templates already lead with a verdict header — don't insert prose above the header.
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.
- 7d ago First seen · 375 lines · 71 tokens per session scan C 75aa5f8f5295
agent-base is a skill published in the GitHub repository volodchenkov/claude-sdlc-agents (2 stars, last pushed 1mo ago), licensed MIT. It adds 71 tokens to every session and 7,353 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
next-partial-prefetching-adoption
Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…
chronicle
Analyze Copilot session history for standup reports, usage tips, session search, and session reindexing. Use when the user asks for a standup, daily summary, usage tips, workflow recommendations, wants to search or find past sessions by keyword/file/PR, wants to reindex their session store, or asks about deleting…
babysit-pr
Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…