fix(bot): improve diagnostic rigor and context awareness

- Mandate the use of `gh run view` for empirical log verification rather than static code inspection.
- Update interactive mode prompt to allow the agent to retain task context and run the unblocking protocol when following up on its own PRs.

tools/gemini-cli-bot/brain/common.md

@@ -103,6 +103,10 @@ safe:
     preserve any existing exemptions (e.g., `-label:security`, `-label:pinned`,
     `-label:"help wanted"`). Never drop existing protections or safety checks
     unless you have proven they are the explicit root cause of the issue.
+3.  **Empirical Log Verification**: Never diagnose a CI, workflow, or test
+    failure based purely on static code inspection. You MUST use the `gh` CLI
+    (e.g., `gh run view --log-failed`) to fetch and read the actual error logs
+    before confirming a hypothesis or proposing a fix.
 
 ## Pull Request Preparation (MANDATORY)
 

tools/gemini-cli-bot/brain/interactive.md

@@ -27,7 +27,12 @@ Before beginning your analysis, you MUST perform the following research:
 2.  **Ignore Pending Tasks**: You are in interactive mode. You MUST explicitly
     ignore any FAILED, STUCK, or pending tasks listed in the
     `lessons-learned.md` Task Ledger. Do not attempt to complete or resume them.
-    Your ONLY goal is to address the user's specific comment.
+    Your ONLY goal is to address the user's specific comment. **EXCEPTION:** If
+    the user's comment is located on a PR authored by you, or relates to an
+    active task in your ledger, you MUST NOT ignore the task context. You must
+    engage the UNBLOCKING PROTOCOL, inspect the PR's current state (including CI
+    failures), and ensure your response addresses the technical reality of the
+    PR.
 3.  **Verify Request Context**: Use the GitHub CLI to verify the current state
     of the issue/PR you were mentioned in. If the user's request is already
     addressed or obsolete, inform them by using the `write_file` tool to save a