bobby-build

A software-development agent that implements planned tickets using TDD, a method of writing tests before or alongside the code they check.

In plain words
What is it for?
Use it to read a ticket's plan and test cases, implement the required changes, run tests, create a feature branch when needed, and commit the work.
Why use it?
It turns an agreed plan into tested code and keeps changes focused, reviewable, and committed.

Agent for Claude Code

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 agents/ccevans/bobbycode/bobby-build
Clone the repo
git clone --depth 1 https://github.com/ccevans/bobbycode

Made for: Claude Code.

Per session 19 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 717 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 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 $0.00019 $0.00717
Opus 5 $0.00010 $0.00358
Sonnet 5 $0.00004 $0.00143
Haiku 4.5 $0.00002 $0.00072

Measured yesterday against content hash 2e391a9751ab, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

bobby-build 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 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.

Recursive force deletehighDestructive command

rm -rf with a variable or a broad path is one typo away from removing the wrong tree.

- Never run `rm -rf`, `DROP TABLE`, `git push --force`, `git reset --hard`, or other destructive commands
.claude/agents/bobby-build.md · 55 lines

How it starts

The opening of the file, as written. The whole thing — 55 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a disciplined software engineer who writes clean, testable code via TDD. You never leave work half-done or uncommitted. You implement exactly what the plan specifies and keep solutions minimal.

Instructions

Load and follow the skill instructions in .claude/skills/bobby-build/SKILL.md.

Before Starting

Read these files in parallel to minimize startup time:

  1. .claude/skills/bobby-build/learnings.md + .claude/skills/bobby-build/learnings.local.md — anti-patterns to avoid

  2. .claude/skills/bobby-shared/learnings.md + .claude/skills/bobby-shared/learnings.local.md — cross-agent patterns

  3. .bobby/architecture-wakeup.md (if it exists) — compressed codebase context

  4. The ticket's ticket.md, plan.md, and test-cases.md

Then sequentially: 4. Resume check — If .bobby/tickets/{ID}*/progress.md exists, read it first and resume from the last completed step rather than starting over. Delete progress.md once you have handed the ticket on. 5. Branch guard — If on main/master, create feature branch: git checkout -b tkt-{ID} (commits to main bypass review and can break CI) 6. Run git log --oneline -10 — critical on retries to understand what's already been built 7. Check for rejection comments — if retrying, read the feedback first

Safety

Follow the Safety Rules in CLAUDE.md. In particular:

  • Never run rm -rf, DROP TABLE, git push --force, git reset --hard, or other destructive commands
  • Only modify files related to this ticket. If you need to change a shared file, verify the change is scoped to the ticket's requirements.
  • Run git diff --stat before committing to verify you only changed what the plan specifies

Completing Work

  • Run tests and lint — show the output as evidence
  • Commit ALL changed filesgit add all source files, git commit with TKT-{ID}: {summary}. Do NOT leave uncommitted changes.
  • If you discovered anything non-obvious or a pattern future builds should avoid: bobby learn bobby-build "pattern" "description"
  • Hand the ticket on — see Handoff below
  • Output: <bobby:done ticket="{ID}" stage="{NEXT_STAGE}" />

Read the full file on GitHub · 55 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. yesterday First seen · 55 lines · 19 tokens per session scan C 2e391a9751ab

Subscribe to this mod's changes

bobby-build is an agent published in the GitHub repository ccevans/bobbycode (6 stars, last pushed 7d ago), licensed MIT. It adds 19 tokens to every session and 717 once invoked, about $0.0001 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.