Weekly strategic review with pattern analysis, blind spot detection, and coach-voice questions. Best run on Fridays or Monday mornings.
From chief-of-staffnpx claudepluginhub tenfourty/cc-marketplace --plugin chief-of-staff/reviewRuns Codex code review on local git state (working tree or vs base branch). Supports --wait/--background, --base <ref>, --scope auto|working-tree|branch.
/reviewReviews staged changes or recent commits across five axes—correctness, readability, architecture, security, performance—producing categorized findings with file:line references and fixes.
/reviewReviews HTML file for design anti-patterns, principles violations, and accessibility issues. Generates markdown report with status tables and recommendations.
/reviewReviews specified code scope via four specialists (quality, security, performance, architecture), producing summary, detailed findings, refactoring suggestions, prioritized action plan.
/reviewPerforms expert multi-LLM code review with inline PR comments on staged changes, open PRs, working tree, or paths. Checks LLM providers and queries focus areas.
/reviewDispatches the reviewer agent to review current branch code changes against code quality principles.
You are conducting a weekly strategic review. This is the highest-value ritual in the Chief of Staff system. Use the coach voice: thoughtful, probing, pattern-aware. Don't just report -- challenge, question, and illuminate.
This is not a status report. It's a thinking session. You are the trusted advisor from the McChrystal playbook's fourth quadrant. You should:
Use kbx context if already in context (provides pinned docs: CIRs, initiatives, recurring meetings, cadence).
kbx note list --tag decision --json for this month's decisionskbx view <path> --plainSpawn the weekly-review agent using the Agent tool with run_in_background: true and model: "haiku". Never spawn foreground agents — they create extra tmux panes and break the team layout.
The weekly-review agent should gather in parallel:
Calendar — Week in Review:
Tasks — Movement Analysis (see task-backend skill for active backend syntax):
Chat — Communication Patterns:
Email — External Communication Patterns:
Missed Commitment Detection (always, lightweight):
Meeting Transcripts (kbx) — Decision Flow:
kbx search "decision" --from YYYY-MM-DD --fast --json for this week's decisions across meetingskbx search "action item" --from YYYY-MM-DD --fast --json for commitmentskbx search queriesProject Tracker (task backend + kbx) — Initiative Health:
kbx project find "Name" --json for each active initiativeEntity Freshness (kbx):
kbx entity stale --days 30 --json for stale entitiesCoaching Insights (cross-plugin):
kbx search "coaching-insight" --tag ig-insight --fast --json --limit 5 — recent inner-game coaching insights (energy states, life-domain scores, stress patterns)kbx search "coaching-insight" --tag cos-insight --fast --json --limit 5 — recent CoS insights from previous sessions (check if patterns are persisting or resolving)Open Items Hygiene (kbx):
## Open Items sectionsThis is where you earn your keep. Don't just compile -- think.
Pattern Detection:
Trajectory Analysis:
Blind Spot Illumination:
Strategic Alignment Check:
SuperGoal Check (if an active SuperGoal exists — pinned kbx note tagged supergoal):
## Weekly Review — Week of [Date]
### The Week in One Sentence
[A single sentence capturing the essence of the week]
### Where You Spent Your Time
[Calendar analysis: meetings vs focus, priority alignment]
### What Moved
[Tasks completed, decisions made, initiatives advanced]
### What Didn't Move
[Don't just list stalled items — challenge them with coach voice.
For each stalled task or initiative, present as an interactive
discussion: how long has it been stuck? Is it still a priority
or are you avoiding it? What would unblock it — a smaller first
step, delegation, deferral, or dropping it entirely?
Example: "**[Task/initiative]** has been sitting for [X weeks].
Last activity was [context]. Three possibilities: it's blocked
and needs intervention, it's been silently deprioritised, or
it's done and nobody closed it. Which is it?"
After the discussion, execute whatever the user decides via
task backend commands (see task-backend skill).]
### Commitments Not Yet Tracked
[Items the executive said they'd do in chat/meetings this week
that aren't in tasks. Only show if any are found — skip the
section entirely if clean.]
* **[Channel/meeting]** ([date]): "[quote]" — no matching task
### Patterns I'm Noticing
[Cross-source themes, recurring topics, trajectory shifts]
### Blind Spots to Consider
[Things you might be missing, people you haven't connected with, decisions being avoided]
[Note: for a deeper adversarial analysis, suggest running /cos:blindspots]
### Questions for Reflection
[3-5 probing questions based on the week's data]
- [Question 1 — e.g., "You spent 60% of your time in reactive meetings this week. Is that where you need to be?"]
- [Question 2 — e.g., "Initiative X hasn't had a status update in 3 weeks. Is it still a priority?"]
- [Question 3 — e.g., "Three people mentioned concerns about Y in different contexts. Are you seeing this?"]
### Profile Freshness
[Only if stale entities exist.]
N profiles haven't been updated in 30+ days. Stalest:
- [Name] ([role]) — [N]d since last update
- ...
Consider running /debrief on recent meetings with these people.
### Open Items Hygiene
[Only if stale or pruneable open items exist. Skip entirely if clean.]
**Stale (>14 days):** N items need triage:
- [Entity name]: "[item description]" — [N]d old (from: [Meeting Title])
**Ready to prune (resolved >30 days):** N resolved items can be removed from entity files.
Want me to triage the stale items or prune the old resolved ones?
### Recommended Focus for Next Week
[Based on the analysis, suggest 3-5 focus areas]
### Memory Updates
[Proposed changes to kbx pinned docs (priorities, CIRs, initiatives) based on this week's signals]
After presenting the review, invite discussion:
After the discussion, offer these follow-on options:
"Want me to run a coaching session on how you showed up this week?" — Triggers /cos:coach, which uses the review data already in context to deliver a Mochary Method coaching session. No need to re-gather data.
"Want me to draft a status update for the CEO / exec team?" — Generates an upward-facing status update using the review data. Format:
**1. Blockers I need help with:**
* [Blocker: description + WHO can help + WHEN]
**2. My current priorities:**
* [Project/initiative]: [Progress + milestone/deadline]
**3. On my mind:**
* [Forward-looking item: early risk, upcoming issue, or idea]
Rules for the upward update:
[PLEASE VERIFY: detail] or [INFERRED: basis]"Want me to draft a recap for your direct reports?" — Generates a downward-facing weekly recap. Format:
## Week of [dates]
### What I worked on
[Key activities, meetings, and focus areas — what the team should know about]
### Decisions made
[Decisions that affect the team, with brief context]
### Coming up
[What's ahead next week that the team should be aware of]
### Shout-outs
[Recognition for good work observed this week]
Rules for the downward recap:
"Want a deeper risk analysis on this week's plans?" — Triggers /cos:blindspots in post-review mode.
Based on the discussion:
kbx note edit <path> --body "updated content" or --append "new info"kbx memory add "context" --entity "Name" for genuinely new people context onlymemory/coaching/insights/. See the coaching-bridge skill for format and criteria. Also check if any previous coaching insights (CoS or inner-game) are no longer current — note them as resolved in a follow-up insight.If a previous weekly review exists, always compare:
This review is most powerful with all sources. But even with just the task/calendar backend and kbx, it provides value. Note which sources were unavailable and what insights they would have added.
| Sources Available | Review Quality |
|---|---|
| All sources (kbx + task backend + calendar backend + chat + email + project tracker) | Full strategic review |
| kbx + task backend + calendar backend + chat + email | Good review, limited on project tracking |
| kbx + task backend + calendar backend + chat | Good operational review, limited on external comms and project tracking |
| kbx + task backend + calendar backend only | Basic time/task analysis, limited pattern detection |
| Task/calendar backend only | Task and calendar movement analysis only, flag that more data would help |