agentic-advisor 1.4.0

Added

• Per-user off switch: LINEARB_AGENTIC_ADVISOR_ENABLED=false (or 0 / off / no) stops the skill from auto-running, so there's no nudge, no hold and nothing reported. Set it in your shell or in ~/.claude/settings.json under env; it works even when the plugin is installed through managed settings. The skill also checks it first, so Claude picking it on its own stops right away. The skill still runs when asked for by name.

Changed

• When the skill graded but didn't print its verdict line, the next file read or edit now gets a one-time reminder instead of being blocked, so there's no red hook error and the call runs on the first try. The first-edit hold (grade before writing code) is unchanged.
• The "no readable transcript" log line now names the hook event and says whether the path was missing or the file wasn't found.

agentic-advisor 1.3.1

Fixed

• The grade now happens before work is handed to a subagent. The first subagent spawn in a repo is held once until the skill runs, like the first edit, and the agent is asked to pass the effort level on in the subagent's prompt. Before, a task that delegated its edits (common in scheduled tasks) was only graded at the main agent's first own edit, after the work was done. Read-only agents (Explore, Plan) are never held.
• Troubleshooting-log fixes from the 1.3.0 review that missed the release: curl's response and error files are now private (0600), emails are redacted before a server response is cut to length, the log no longer records the ticket key from the prompt, and the version prefix no longer needs jq.

agentic-advisor 1.3.0

Changed

• The troubleshooting log is now on by default, so a missing event can be diagnosed without asking the user to re-run anything. Turn it off with LINEARB_DEBUG=0. It stays local and private (0600), rotates at 1 MB, and never contains the token, prompt or payload.
• The log now covers the trigger hook too: when it nudged the agent to run the skill, and when it held a tool call for the grade or the verdict line.
• Every line starts with the plugin version.

Added

• Errors are logged with their reason: network failures with curl's exit code and message, and rejected events (4xx/5xx) with the server's response, emails redacted.

agentic-advisor 1.2.0

Added

• Ticket context: when the prompt or branch names a ticket (e.g. ABC-123), the skill fetches its type, priority, estimate and labels from LinearB (one call, same token) and uses them to raise the task-complexity grade, e.g. a Highest-priority bug. Skipped silently when there's no key or no PM integration. (#10)

agentic-advisor 1.1.2

Added

• Opt-in troubleshooting log: LINEARB_DEBUG=1 writes each report's outcome (skip reason or HTTP status) to ${XDG_CACHE_HOME:-~/.cache}/agentic-advisor/report.log, rotated at 1 MB. Lines for sent events include the grade and repo name; no line contains the token, email, prompt or payload. The log stays private (folder 0700, file 0600, symlinked logs and rotation targets refused). (#9)

Changed

• With no LINEARB_API_TOKEN, the verdict now says LINEARB_API_TOKEN not set and tells the user how to set it, instead of a generic "signals unavailable" that looked like a working setup. (#9)
• Hooks no longer fail under set -u when HOME and XDG_CACHE_HOME are both unset. (#9)