From 4e8c018f830f1448c7ac2410b042a950a597c2e6 Mon Sep 17 00:00:00 2001 From: "david.delagneau" Date: Thu, 30 Apr 2026 15:05:49 -0600 Subject: [PATCH] feat: mandate real-time auto-documentation of project memory updates across all interaction turns --- GEMINI.md | 16 ++++++++++++---- agent-memory/behavior/agent-behavior.md | 6 +++--- 2 files changed, 15 insertions(+), 7 deletions(-) diff --git a/GEMINI.md b/GEMINI.md index cf6ac9f..75e8914 100644 --- a/GEMINI.md +++ b/GEMINI.md @@ -22,8 +22,16 @@ Read these first: - `project-knowledge/` is canonical project memory. - `agent-memory/` is agent operating memory. -## Gemini-Specific Notes +## Gemini-Specific Mandatory Loop (Real-Time Auto-Documentation) -- Prefer the shared workspace rules over model improvisation. -- Before answering current-state questions, inspect the current-work files and latest relevant daily note. -- If a prompt changes durable project understanding or reusable agent behavior, update the smallest correct canonical file. +At the end of EVERY conversational turn, you MUST evaluate the interaction as a senior Obsidian documentation expert: + +1. **Analyze for Memory Impact:** Did the user's prompt or your subsequent action reveal new technical context, resolve an open question, change an architectural assumption, or shift current priorities? +2. **Identify Target Canonical Files:** If memory impact is detected, identify WHERE in the `project-knowledge/` directory this information belongs. This is NOT limited to daily logs. It could be: + - `01-current/current-work.md` (priority shifts, new blockers) + - `02-work-items/*.md` (story updates, bug reproduction steps) + - `03-context/` (system architecture, processes, workflows) + - `04-people/` (stakeholder involvement) + - `06-daily/YYYY-MM-DD.md` (daily findings) +3. **Proactive Update:** You MUST use your file editing tools (`replace_file_content` / `multi_replace_file_content`) to inject the new context or CORRECT existing stale context in the appropriate files BEFORE concluding your response. +4. **No Explicit Prompt Needed:** Do not wait for the user to say "document this". Auto-document implicitly. Ensure you do not pollute the workspace with unverified guesses; only document confirmed context or active hypotheses. diff --git a/agent-memory/behavior/agent-behavior.md b/agent-memory/behavior/agent-behavior.md index f548972..8ef0df4 100644 --- a/agent-memory/behavior/agent-behavior.md +++ b/agent-memory/behavior/agent-behavior.md @@ -13,9 +13,9 @@ Make workspace memory maintenance part of the agent's normal job, not a separate --- -## Default Agent Behavior +## Default Agent Behavior (Real-Time Auto-Documentation) -For any meaningful prompt, the agent should decide whether the interaction changes project memory. +For EVERY meaningful prompt, the agent MUST execute a real-time auto-documentation loop. It is not an optional workflow; it is the core of the agent's job. This applies to: @@ -27,7 +27,7 @@ This applies to: - debugging discussions - corrections to previous understanding -The agent should not wait for a separate promotion command when the right update is already clear. +The agent MUST NOT wait for a separate promotion command. The agent MUST proactively use file editing tools to update `project-knowledge/` (including 01-current, 02-work-items, 03-context, 04-people, 06-daily) within the SAME conversational turn if new information is detected or if prior context becomes stale. - When editing `project-knowledge/`, write as a human engineer maintaining shared project documentation. - Keep agent-operating logic out of `project-knowledge/`; store that logic in prompts, commands, skills, agents, or `agent-memory/`.