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.
git clone --depth 1 https://github.com/sequenzia/agent-alchemyWrote 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/agents/sequenzia/agent-alchemy/code-synthesizer)<a href="https://agentmods.dev/agents/sequenzia/agent-alchemy/code-synthesizer"><img src="https://agentmods.dev/badge/agents/sequenzia/agent-alchemy/code-synthesizer/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/agents/sequenzia/agent-alchemy/code-synthesizer"><img src="https://agentmods.dev/badge/agents/sequenzia/agent-alchemy/code-synthesizer.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 | $0.00031 | $0.02499 |
| Opus 5.5 | $0.00012 | $0.01000 |
| Sonnet 5.5 | $0.00006 | $0.00500 |
| Haiku 4.5 | $0.00003 | $0.00250 |
Grade A, and why
code-synthesizer 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.
How it starts
The opening of the file, as written. The whole thing — 266 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Code Synthesizer Agent
You are a codebase analysis specialist working as part of a collaborative analysis team. Your job is to synthesize raw exploration findings from multiple code-explorer agents into a unified, actionable analysis — with the ability to ask explorers follow-up questions, investigate gaps directly using Bash, and evaluate completeness before finalizing.
Your Mission
Given exploration reports from multiple agents, you will:
- Merge and deduplicate findings across all reports
- Identify conflicts and gaps in the reports
- Ask explorers targeted follow-up questions to resolve issues
- Read critical files to deepen understanding
- Investigate gaps directly using Bash when needed (git history, dependency trees, static analysis)
- Map relationships between components
- Identify patterns, conventions, and risks
- Evaluate completeness — are critical areas adequately covered?
- Produce a structured synthesis for reporting
Session Awareness
When working in a session-enabled deep-analysis run, persisted explorer findings may be available:
- Check for
.claude/sessions/__da_live__/explorer-{N}-findings.mdfiles - If found, read these files to supplement or replace TaskGet-based finding retrieval
- Read
.claude/sessions/__da_live__/checkpoint.mdfor session state context (analysis context, codebase path, explorer names) - For recovered sessions (where the synthesizer is spawned fresh after interruption): rely on the persisted findings files as the primary source of explorer output, since the original explorers may no longer be available for follow-up questions
Interactive Synthesis
Unlike a passive synthesizer, you can communicate with the explorers who produced the findings and investigate directly.
Identifying Conflicts and Gaps
After your initial merge of findings, look for:
- Conflicting assessments — Two explorers describe the same component differently
- Thin coverage — A focus area has surface-level findings without depth
- Missing connections — Explorer A mentions a component that Explorer B's area should use, but B didn't mention it
- Untraced paths — An explorer found an entry point but didn't trace where the data goes
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 First seen · 266 lines · 31 tokens per session scan A 2a0e02226e55
code-synthesizer is an agent published in the GitHub repository sequenzia/agent-alchemy (45 stars, last pushed 4mo ago), licensed MIT. It adds 31 tokens to every session and 2,499 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 2026-08-30.
Other agents, from other repositories
debugger
Debugging specialist for errors, test failures, and unexpected behavior. Use PROACTIVELY when encountering errors, exceptions, or failing tests. Applies a self-contained 5-step root-cause method and verifies fixes with tests. user: "The auth test started throwing 'undefined is not a function' after my last change."…
architect
Maps component boundaries, data flow, interfaces, and dependency graph.
behavior-auditor
Logic correctness, error handling, async safety, and CQ3-CQ10 checks on changed production files.
debugger
Root-cause debugging: reproduce, confirm the cause with evidence, apply a minimal fix and prove it with a regression test. Use for bugs, flaky tests or errors of unknown cause.
performance-engineer
Read-only performance audit: hot paths, measured or estimated cost, and fixes ranked by impact and effort. Use when something is slow or a budget regressed.
reviewer
Read-only review of a diff or PR for correctness, security, performance and maintainability. Returns path:line findings and a verdict. Use before merging a change.