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 agentmods add skills/fortiumpartners/ensemble/git-townnpx skills add FortiumPartners/ensemble --skill git-towngit clone --depth 1 https://github.com/FortiumPartners/ensembleWrote 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/fortiumpartners/ensemble/git-town)<a href="https://agentmods.dev/skills/fortiumpartners/ensemble/git-town"><img src="https://agentmods.dev/badge/skills/fortiumpartners/ensemble/git-town.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 | $0.00127 | $0.05075 |
| Opus 5 | $0.00063 | $0.02537 |
| Sonnet 5 | $0.00025 | $0.01015 |
| Haiku 4.5 | $0.00013 | $0.00507 |
Grade A, and why
git-town 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 yesterday.
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.
curl -L https://github.com/git-town/git-town/releases/download/v14.0.0/git-town_linux_amd64 -o /usr/local/bin/git-town How it starts
The opening of the file, as written. The whole thing — 654 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Git-Town Workflow Skill
Skill Type: Technical Workflow Estimated Learning Time: 30 minutes Proficiency Levels: Beginner, Intermediate, Advanced
Mission
Provide comprehensive git-town workflow integration for Claude Code agents, enabling autonomous branch management, PR creation, and error recovery without interactive prompts. This skill empowers agents to execute complex git workflows with zero user intervention by leveraging git-town's CLI flags for non-interactive operation. Target audience includes orchestrator agents (tech-lead-orchestrator, git-workflow) and developer agents (frontend-developer, backend-developer) performing feature development, bug fixes, and release workflows. Key capabilities include validation-first execution with structured exit codes, standardized error handling patterns for merge conflicts and remote failures, stacked branch workflows for complex features, offline development mode support, and GitHub CLI integration for automated PR creation. Agents using this skill can autonomously manage the complete feature lifecycle from branch creation through PR merge without requiring user intervention for routine git operations.
Skill Loading Mechanism
Discovery Paths
Git-town skill files are located using XDG Base Directory Specification with the following search priority:
- Primary:
$XDG_CONFIG_HOME/ensemble/skills/git-town/(user-specific configuration) - Fallback:
~/.config/ensemble/skills/git-town/(default XDG location) - Legacy:
~/.ensemble/skills/git-town/(backward compatibility) - Plugin:
<plugin-root>/packages/git/skills/git-town/(bundled with ensemble-git plugin)
Agents should check paths in order and use the first match. Plugin-bundled files serve as the fallback when user configuration is absent.
Loading Performance
Git-town skill loading is optimized for agent efficiency with the following performance characteristics:
- Target: <100ms for full skill load (SKILL.md + REFERENCE.md + ERROR_HANDLING.md combined)
- Caching: Skill content is cached in agent context memory after first load, with cache invalidation triggered by file modification timestamps (mtime comparison)
- Lazy loading: Templates and migration guides are loaded on-demand only when agents query specific sections, reducing initial memory footprint
- Indexing: Section headers are indexed during initial load using hash maps for O(1) lookup of subsections, enabling fast targeted queries
What ships with it
24 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- .claude/settings.local.json 65 B
- cicd/examples/github-actions.yml 20 KB
- cicd/examples/gitlab-ci.yml 16 KB
- cicd/INTEGRATION.md 22 KB
- COMPLETION_SUMMARY.md 16 KB
- ERROR_HANDLING.md 94 KB
- guides/agent-skill-integration.md 20 KB
- guides/migration-git-flow.md 21 KB
- guides/migration-github-flow.md 20 KB
- guides/migration-trunk-based.md 22 KB
- guides/monorepo.md 21 KB
- guides/onboarding.md 9.8 KB
- README.md 18 KB
- REFERENCE.md 13 KB
- scripts/test-validate.sh 7.5 KB runs code
- scripts/validate-git-town.sh 7.0 KB runs code
- templates/interview-branch-creation.md 11 KB
- templates/interview-completion.md 15 KB
- templates/interview-pr-creation.md 15 KB
- templates/SKILL_TEMPLATE.md 12 KB
- TESTING.md 19 KB
- tests/comprehensive-test.sh 12 KB runs code
- tests/test-integration.sh 14 KB runs code
- tests/test-skill.sh 11 KB runs code
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.
- yesterday First seen · 654 lines · 127 tokens per session scan A 2f7234046b5b
git-town is a skill published in the GitHub repository FortiumPartners/ensemble (11 stars, last pushed 1mo ago), licensed MIT. It adds 127 tokens to every session and 5,075 once invoked, about $0.0006 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 2026-09-03.
Other skills, from other repositories
saas-payments
Implement production-ready payment processing with Stripe including one-time payments, recurring subscriptions, customer portal, metered billing, and webhook handling. This skill covers the complete billing lifecycle from checkout through subscription management.
saas-auth
Implement production-ready authentication for SaaS applications including email/password, OAuth social login, session management, and security best practices. This skill covers the complete auth lifecycle from signup through password recovery.
saas-analytics
Implement product analytics for understanding user behavior, tracking key metrics, and making data-driven decisions. Covers event tracking, user identification, cohort analysis, and integration with analytics platforms like PostHog, Mixpanel, or custom solutions.
saas-onboarding
Implement effective user onboarding flows that drive activation and reduce churn. Covers setup wizards, progress checklists, contextual tooltips, and tracking of key activation milestones. Focus on getting users to their "aha moment" quickly.
saas-billing
Implement subscription lifecycle management, plan enforcement, usage tracking, and billing operations. Covers trial periods, plan changes, quota enforcement, and subscription status synchronization with payment providers.
saas-email
Implement reliable transactional email delivery for SaaS applications including templating, queuing, delivery tracking, and provider abstraction. Covers common email types: welcome, verification, password reset, notifications, and invoices.