Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/varienos/agentic-workflownpx agentmods add commands/varienos/agentic-workflow/bootstrapWrote 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/commands/varienos/agentic-workflow/bootstrap)<a href="https://agentmods.dev/commands/varienos/agentic-workflow/bootstrap"><img src="https://agentmods.dev/badge/commands/varienos/agentic-workflow/bootstrap/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.
<a href="https://agentmods.dev/commands/varienos/agentic-workflow/bootstrap"><img src="https://agentmods.dev/badge/commands/varienos/agentic-workflow/bootstrap.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00000 | $0.37124 |
| Opus 5.5 | $0.00000 | $0.14850 |
| Sonnet 5.5 | $0.00000 | $0.07425 |
| Haiku 4.5 | $0.00000 | $0.03712 |
Grade E, and why
bootstrap scanned grade E with 4 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 22d 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.
Downloads and executes remote codehighSupply chain
curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.
curl -LsSf https://astral.sh/uv/install.sh | sh Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
test -f ./.claude/settings.json && echo "✅ C1: settings.json present (claude_runtime_dosyalari_var)" || echo "❌ C1: settings.json MISSING" Reads MCP configurationmediumAgent snooping
mcp.json carries server URLs and auth tokens; reading it lets a mod discover and abuse other integrations.
grep -q '"basic-memory"' ./.mcp.json && echo "✅ I2: .mcp.json basic-memory entry present" || echo "❌ I2: .mcp.json basic-memory entry MISSING" Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -LsSf https://astral.sh/uv/install.sh | sh How it starts
The opening of the file, as written. The whole thing — 2,948 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Bootstrap — Agentic Workflow Setup Wizard
This command analyzes your project, runs a short interview with you, and creates the full agentic workflow configuration. Every step is visible on the console. When finished, the .claude/ directory, root files, manifest, and backlog are ready.
RULE: Follow these instructions literally, STEP BY STEP, in order. Skip no step, merge no steps. Show the user the outputs from every step.
Invariant rules
These rules are the foundation of Bootstrap and of every file it produces:
1. Git runs only in Codebase
- There is NO
.git/inside Agentbase. Agentbase is a configuration directory only. - All git operations (commit, push, branch, worktree) run inside
../Codebase/. - Git commands in commands:
cd ../Codebase && git ... - Worktree isolation:
cd ../Codebase && git worktree add ... - This split keeps Codebase's safe worktree isolation independent of agent and backlog files.
- Two-repo delivery model (optional): The project root (the parent ABOVE Agentbase/Codebase/Docbase) may be the developer's OWN git repo — it versions Agentbase + Docbase and ignores
Codebasevia.gitignore(see STEP 6.6). Codebase is its own independent repo and is delivered SEPARATELY to the customer (the customer clones only Codebase). Even in that case.gitis written at the project root; it is NOT written inside Agentbase — the "no.git/in Agentbase" rule above still holds. - Agent boundary: Agent/workflow git operations (commit/push/branch/worktree) ALWAYS target
../Codebase/; the agent never touches the parent (developer) repository. The parent-root repo is the developer's manual tool, not the agents'.
2. Bootstrap NEVER writes to Codebase (single exception: AI Import)
- Bootstrap READS Codebase → CONFIGURES Agentbase.
- No file in Codebase is changed, added, or deleted.
- All generated files go into Agentbase/ — distribution:
- Agentbase ROOT (i.e. directly under
Agentbase/) — produced DIRECTLY by Bootstrap (6+1 root documents + 1 mcp config):PROJECT.md,STACK.md,DEVELOPER.md,ARCHITECTURE.md,WORKFLOWS.md,CLAUDE.md(root context),onboarding.md(new developer guide),.claude-ignore,.mcp.json(required — fromtemplates/core/mcp.skeleton.json). Produced by tools Bootstrap invokes:backlog/(Backlog.md CLI),../Docbase/memory/(basic-memory vault). Shipped with the repo (static root documents — Bootstrap does not fill these; they are included via the@<file>chain in rootCLAUDE.md):ORCHESTRATION.md(shared agent behavior philosophy — for all models),LESSONS.md(self-improvement lessons),BACKLOG.md(Backlog CLI quick reference). Shipped with the repo (code and template infrastructure):bin/,templates/,tests/,generate.js,transform.js,package.json. - Under Agentbase/.claude/:
commands/,agents/,hooks/,rules/,reports/,tracking/,custom/,settings.json,CLAUDE.md(internal runtime config for the agent — a SEPARATE file from rootCLAUDE.md, aimed at the agent, not the end user). - Manifest:
../Docbase/agentic/project-manifest.yaml(outside Agentbase, under Docbase). - Optional transform.js outputs (when
manifest.targetshas a target other thanclaude):GEMINI.md(gemini/antigravity target → Agentbase root),.agents/workflows/*,.agents/skills/*/SKILL.md,.agents/rules/*(antigravity target),AGENTS.md+.codex/skills/*/SKILL.md(codex target → Agentbase root +.codex/),.kimi/skills/,.kimi/agents/(kimi target),.opencode/AGENTS.md+.opencode/skills/+.opencode/agents/(opencode target). These files convert rootCLAUDE.mdcontent into the target CLI format — the injection chain is preserved automatically. - FORBIDDEN: Do NOT WRITE or COPY root documents (PROJECT.md, STACK.md, DEVELOPER.md, ARCHITECTURE.md, WORKFLOWS.md, ORCHESTRATION.md, LESSONS.md, BACKLOG.md, root CLAUDE.md, onboarding.md) under
.claude/..claude/is agent runtime config, not documentation. These files stay at the real Agentbase root so all models (Claude, Gemini, Antigravity, Codex, Kimi, OpenCode) can read the same context.
- Agentbase ROOT (i.e. directly under
- The manifest goes under
../Docbase/agentic/(outside Codebase). - The project's existing .gitignore, package.json, and CI config files are preserved.
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.
- 22d ago Changed · -4 lines 8e541f6d7306
- 1mo ago First seen · 2,952 lines · 0 tokens per session scan E 4ae90e1d48ac
bootstrap is a command published in the GitHub repository varienos/agentic-workflow (59 stars, last pushed 22d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 37,124 tokens. A static security scan graded it E with 4 findings (downloads and executes remote code, reads agent configuration directories, reads mcp configuration). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
review-pr
PR review using parallel specialized agents for code quality, security, testing, architecture, and performance analysis. Synthesizes findings into a review report with conventional comments (praise/issue/suggestion/nitpick) and approve or request-changes verdict. Use when reviewing pull requests, conducting security…
review
A command for reviewing code with security checks and a check for compliance with MX tags. The available description does not explain what MX tags are.
gh-security-review
Use to perform a dedicated security review of current branch changes before pushing.
revisar
Peer review estricto de código con análisis de rendimiento, arquitectura, seguridad y cumplimiento. Usa cuando el usuario dice "revisar código", "code review", "peer review", "revisame este archivo", "review my code", "analizar código", "revisar PR", "revisar feature". Incluye análisis de performance, security OWASP…
review-decide
Resolve a circuit breaker in the review loop. Accepts exactly one argument: approve, reject, or retry.
dev-assess
Spawn 3 senior developer subagents in parallel, each assessing the target from a different engineering perspective. Experts are dynamically selected based on the target material's technical domains. Triangulate their findings into an interactive HTML artifact with convergence/divergence analysis.