From e1f97ba1cd5047d6490177f698ea13f2bd416950 Mon Sep 17 00:00:00 2001 From: artale Date: Mon, 29 Jun 2026 12:24:10 +0200 Subject: [PATCH] docs(tac): add autonomous work agent guidance Capture the NanoClaw autonomous work-agent blueprint in the TAC loop-engineering meta-prompt. Require plans to distinguish shared agent factories from personal work agents, isolate each agent, separate credentials through vault/proxy injection, define access policies, and include management/revocation surfaces. Tests lock the new markers. --- pipeline/test_loop_engineering_meta_prompt.py | 4 ++++ plans/meta-prompts/loop_engineering.md | 19 +++++++++++++++++++ 2 files changed, 23 insertions(+) diff --git a/pipeline/test_loop_engineering_meta_prompt.py b/pipeline/test_loop_engineering_meta_prompt.py index becd30e..95f2751 100644 --- a/pipeline/test_loop_engineering_meta_prompt.py +++ b/pipeline/test_loop_engineering_meta_prompt.py @@ -19,6 +19,10 @@ def test_loop_engineering_meta_prompt_locks_safety_fields(): "Open-source model portability", "quality gate checklist", "YouTube `dJI2GRG1GEE`", + "Autonomous work agent", + "Team manages agents", + "vault/proxy credential injection", + "YouTube `hLUGXO5DSpo`", "Model Workspace Protocol", "00-intake/", "40-ship/", diff --git a/plans/meta-prompts/loop_engineering.md b/plans/meta-prompts/loop_engineering.md index 056f0a0..2b975c8 100644 --- a/plans/meta-prompts/loop_engineering.md +++ b/plans/meta-prompts/loop_engineering.md @@ -66,6 +66,25 @@ For outbound/productized work, require: Reference: Jordan Urbs, "GLM 5.2 Proves Open Source AI Can Match Fable 5 (AI Harness Engineering)" (YouTube `dJI2GRG1GEE`). +## Autonomous work agent note + +For business adoption, distinguish two operating modes before designing the harness: + +1. **Team manages agents** — an agent factory or admin team creates, updates, and governs shared agents. +2. **Agents assist people** — each human gets a personal work agent that learns their context, skills, preferences, and second-brain material. + +Every autonomous work-agent plan must include: + +- per-agent isolation boundary, preferably one container or equivalent sandbox per agent; +- credential separation: no raw secrets inside the agent environment; +- vault/proxy credential injection for only the requests the agent is allowed to make; +- explicit access policies for tools, files, APIs, and outbound network; +- skill/instruction iteration loop for tuning the agent's actual output; +- management surface for upgrading, maintaining, and revoking agents after deployment; +- second-brain/context source plan when personal-agent behavior depends on private knowledge. + +Reference: Latent Space, "The Blueprint for Autonomous Work Agents" with Gavriel Cohen / NanoClaw (YouTube `hLUGXO5DSpo`). + ## Model Workspace Protocol note For sequential workflows with human review between stages, prefer folder-structured orchestration before multi-agent framework code. The Model Workspace Protocol pattern treats numbered folders as stages, markdown files as role/context carriers, and local scripts as the boring mechanical layer.