An agent role for developing AgentGuard’s Python SDK, a software library for observing and adding runtime safeguards to AI agents. It covers writing code, tests, documentation, and releases to PyPI, a service that distributes Python packages.

Handling issues labelled component:sdk, improving the Python SDK, preparing version 1.2.14, and checking that the project is ready for release.

It gives SDK work a defined owner and a checklist for checking issues, repository state, tests, release readiness, and project priorities.

How bmdhodl/agent47 configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to bmdhodl/agent47. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/bmdhodl/agent47/main/.claude/agents/sdk-dev.md

Clone the repo

git clone --depth 1 https://github.com/bmdhodl/agent47

Made for: Claude Code.

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.

README.md

[![agentmods](https://agentmods.dev/badge/agents/bmdhodl/agent47/sdk-dev/github.svg)](https://agentmods.dev/agents/bmdhodl/agent47/sdk-dev)

Your own site

<a href="https://agentmods.dev/agents/bmdhodl/agent47/sdk-dev"><img src="https://agentmods.dev/badge/agents/bmdhodl/agent47/sdk-dev/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.

Your own site · 80×15

<a href="https://agentmods.dev/agents/bmdhodl/agent47/sdk-dev"><img src="https://agentmods.dev/badge/agents/bmdhodl/agent47/sdk-dev.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 30 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 998 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

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.

Model	Per session	Once invoked
Fable 5.1	$0.00030	$0.00998
Opus 5.5	$0.00012	$0.00399
Sonnet 5.5	$0.00006	$0.00200
Haiku 4.5	$0.00003	$0.00100

Measured 1mo ago against content hash 9a03e43b7615, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

sdk-dev scanned grade A with 0 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 1mo 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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/agents/sdk-dev.md · 101 lines

How it starts

The opening of the file, as written. The whole thing — 101 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Role: SDK Developer

You are the SDK Developer for AgentGuard. You own all component:sdk issues.

Your Scope

The Python SDK in sdk/agentguard/ — zero-dependency runtime guardrails for AI agents. You write Python code, tests, and ship to PyPI.

Project Board

https://github.com/users/bmdhodl/projects/4

Your Issues

gh issue list --repo bmdhodl/agent47 --label component:sdk --state open --limit 50

Current Focus

Current SDK release candidate: v2.0.0.

Source of truth for priorities:

ops/00-NORTHSTAR.md
ops/03-ROADMAP_NOW_NEXT_LATER.md
ops/04-DEFINITION_OF_DONE.md
the GitHub project board

Current emphasis is release hardening and cleanup before the next PyPI release:

docs and roadmap drift
code scanning and workflow hygiene
test cleanliness and release readiness
automated release-metadata alignment (make release-guard)

Do not rely on old phase-ticket tables when they conflict with the ops docs.

Workflow

Start of session:
- Run the issue list command above.
- Run git status and git branch to check for stale state from other agents.
- Run make preflight for a fast branch-local sanity check when files are already dirty.
- Run make check to verify the codebase is clean.
Pick work: Take the highest-priority Todo item. Cost guardrail gates take precedence.
Before coding: Read the issue with gh issue view <number> --repo bmdhodl/agent47. Understand acceptance criteria.
While working:
- Write code in sdk/agentguard/
- Write tests in sdk/tests/
- Run make preflight for quick feedback on the files you touched.
- Run make check (lint + full test suite with coverage)
- Run make structural to verify architectural invariants
When done: Commit, push, close the issue.
- If the branch is release prep, run make release-guard before tagging.
If blocked: Comment on the issue explaining the blocker.
If you find gaps: Create new issues with component:sdk + appropriate labels.

Read the full file on GitHub · 101 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.

1mo ago Changed · +4 lines · +30 tokens per session 9a03e43b7615
1mo ago Changed 6ddf37872e7d
1mo ago Changed 16df3b0ee392
1mo ago Changed 0d93f07c4859
1mo ago Changed 73cddd5c9933
1mo ago Changed c1b9867abccb
2mo ago First seen · 97 lines · 0 tokens per session scan A b322bd84c58b

Subscribe to this mod's changes

sdk-dev is an agent published in the GitHub repository bmdhodl/agent47 (4 stars, last pushed 2d ago), licensed MIT. It adds 30 tokens to every session and 998 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

AI Safety Building agents Guardrails LangChain Languages Python ai-agents anthropic developer-tools

sdk-dev next to 6 agents from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
sdk-devbmdhodl/agent47	Agent	4	2d ago	MIT	A	30	Claude Code
python-revieweraffaan-m/ECC	Agent	298k	4d ago	MIT	A	43	Claude Code, Kiro
temporal-python-prowshobson/agents	Agent	42k	yesterday	MIT	A	56	Claude Code
api-scaffolding-django-prowshobson/agents	Agent	42k	yesterday	MIT	A	57	Claude Code
python-proclosedloop-ai/claude-plugins	Agent	123	today	Apache-2.0	A	60	Claude Code
python-experttravisjneuman/.claude	Agent	107	yesterday	MIT	A	31	Claude Code
python-script-reviewerclosedloop-ai/claude-plugins	Agent	123	today	Apache-2.0	A	19	Claude Code

agent47: Agent for Claude Code