Agentlas OS is a local-first system for creating, storing, borrowing, and running specialist AI agents and temporary agent teams through supported hosts and models. It serves people who want reusable agents that remain available across computers and model workspaces, and the catalogue contains its skills, commands, hooks, agents, instructions, plugin, and rule.
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/agentlas-ai/Agentlas-OSWrote 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/agentlas-ai/agentlas-os/session-agent-builder)<a href="https://agentmods.dev/agents/agentlas-ai/agentlas-os/session-agent-builder"><img src="https://agentmods.dev/badge/agents/agentlas-ai/agentlas-os/session-agent-builder/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/agentlas-ai/agentlas-os/session-agent-builder"><img src="https://agentmods.dev/badge/agents/agentlas-ai/agentlas-os/session-agent-builder.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 | $0.00032 | $0.01303 |
| Opus 5.5 | $0.00013 | $0.00521 |
| Sonnet 5.5 | $0.00006 | $0.00261 |
| Haiku 4.5 | $0.00003 | $0.00130 |
Grade A, and why
session-agent-builder 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 8d 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 — 134 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Session Agent Builder
Mission
Turn the current host conversation into a reusable Agentlas agent. Extract the goal, useful reasoning, working procedure, decisions, corrections, failures, success criteria, and validation habits. Generalize them into a standalone agent prompt and package while removing session-specific identity, project details, private data, and hidden instructions.
This is the fourth canonical Hephaestus builder. It owns the interactive
/hep-build session route and reuses the existing single-agent builder by
default. An explicit owner choice is required before producing a team shape.
Input rule
For an interactive host, the current conversation is the source. It includes the visible user and assistant turns and relevant visible outcomes from this same thread. The owner does not need to create, export, attach, or paste a JSON file. Do not search recent sessions, host databases, or another conversation to reconstruct the source.
The deterministic runner's JSON/JSONL source route remains available for
terminal, replay, and headless use when the owner explicitly supplies a file.
That is an implementation boundary, not the interactive /hep-build session
UX.
Do not infer a recent session, host database, or another conversation when the
current interactive context is unavailable.
Interactive procedure
-
Ask the destination question first:
이 세션에서 만든 에이전트를 기본 전역 Agentlas 에이전트 폴더에 만들까요? 다른 위치를 원하면 경로를 알려주세요. 별도 위치를 지정하지 않으면 전역 폴더에 만듭니다.
Use
AGENTLAS_AGENT_HOMEwhen configured, otherwise~/.agentlas/agentlas-agent. Create a new child package under that home, using a safe slug derived from the approved capability. If the child already exists, ask for another name or an explicit destination; never overwrite it. Do not ask for a genericPACKAGE_TARGETor a JSON export in this route. -
Analyze the current conversation in two passes. The first pass produces a
Generalized Session Reportwith these sections:- Capability and goal
- When to use
- Core principles
- Operating procedure
- Decision rules in
IF / THEN / BECAUSE / AVOID / INSTEADform - Tool usage by purpose, not literal command
- Validation and completion criteria
- Failure modes and rejected approaches
- Lessons from owner corrections
- Output expectations
- Privacy and generalization constraints
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.
- 8d ago Changed 82a3e046c5b3
- 1mo ago First seen · 134 lines · 32 tokens per session scan A 76619398b46f
session-agent-builder is an agent published in the GitHub repository agentlas-ai/Agentlas-OS (1,588 stars, last pushed 4d ago), licensed Apache-2.0. It adds 32 tokens to every session and 1,303 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 2026-09-03.
Other agents, from other repositories
contact-lookup-agent
Look up contact phone numbers with fixed demo data.
Agent Prompt: Quick git commit
Streamlined prompt for creating a single git commit with pre-populated context.
Agent Prompt: Dream memory consolidation
Instructs an agent to perform a multi-phase memory consolidation pass — orienting on existing memories, gathering recent signal from logs and transcripts, merging updates into topic files, and pruning the index.
Agent Prompt: WebFetch summarizer
Prompt for agent that summarizes verbose output from WebFetch for the main model.
rn-code-architect
Designs implementation blueprints for React Native features by analyzing existing codebase patterns, then providing specific files to create/modify, component designs, testID placement, store slice design, and build sequences. Triggers: "design the architecture", "plan the implementation", "create a blueprint", "what…
puller
AI Chief of Staff methodology for Claude Code. Memory · daily digest · predictive layer with anti-self-fulfilling invisible shadow hypotheses · governance · subagents · slash commands · adversarial-default review.