From 26f868b39973e8debe134f0bbaa343a05a0e8464 Mon Sep 17 00:00:00 2001 From: Levi Neely <141506390+lneely@users.noreply.github.com> Date: Fri, 1 May 2026 12:53:37 +0200 Subject: [PATCH] Update TRIAL_1.md --- experiments/TRIAL_1.md | 13 +++++-------- 1 file changed, 5 insertions(+), 8 deletions(-) diff --git a/experiments/TRIAL_1.md b/experiments/TRIAL_1.md index 91ab1c4..181f704 100644 --- a/experiments/TRIAL_1.md +++ b/experiments/TRIAL_1.md @@ -102,14 +102,11 @@ The following transitions required manual user intervention (the user prompting ## Key Observations & Behavioral Patterns -- **Compliance vs. Correction**: A critical failure point was identified where the developer agent prioritized "Prompt Compliance" (creating `CMakeLists.txt` as instructed) over "System Accuracy" (noting that the project doesn't use CMake). This suggests that agents in a specialized role may be less likely to challenge the parent's premises. -- **Reviewer Myopia**: The build reviewer successfully identified missing flags in the `Makefile` but initially ignored the presence of the hallucinated CMake instruction. Reviewers appear to be effective at checking "correctness" of changes but less effective at identifying "extraness" or "repository pollution" unless specifically tasked with "purity." -- **The "Done" State Trap**: Agents frequently entered a "logical completion" state where they believed the task was over but failed to perform the "administrative completion" of notifying the parent. This indicates that the transition from *thinking* to *system-signaling* is the weakest link in the P2P choreography. -- **Information Decay in Parallelism**: In a parallel spawn, context must be perfectly mirrored. Any discrepancy in the roster or instructions between concurrent agents can lead to deadlocks where one agent waits for a signal that another agent doesn't know it was supposed to send. - -## Conclusion -The model is effective for multi-stage engineering tasks. However, notification protocols rely on agent autonomy which sometimes fails. Manual inspection of session files remains a critical fallback for the orchestrator. +- ollie enables concurrent multi-agent development workflows. +- Agent autonomy and information decay may have limited how effectively agents coordinated amongst themselves +- Parent agent accurately assesses current state by mass inspection of all sub-agents via filesystem +- Accurate resumption of work possible with minimal human intervention ## Future Work -- **Just-in-Time (JIT) Agent Generation**: Ollie can dynamically generate specialized JSON agent configs in `a/`, tailored to the immediate task (e.g., a specific kind of reviewer such as `pcloudccc-security-reviewer`). This approach enables ollie to inject specialized hooks or system prompts that, for example, provide the current environment / roster every turn, or keep the coordination instructions fresh in the context. +- **Just-in-Time (JIT) Agent Generation**: The information decay / agent autonomy hypothesis can be tested. ollie agents can inject prompts, either as part of the system prompt or as part of hook execution. ollie can therefore dynamically generate specialized JSON agent configs for a given task, and store them in `a/`. Any agent in `a/` is a valid argument for the `agent` option in `subagent_spawn` (e.g., `subagent_spawn ... -agent=my-jit-agent ...`).