Agent Skills: Status Updates

Status Updates

Expert guidance for crafting project updates, managing stakeholder communications, documenting decisions, surfacing risks, and running productive meetings across product and engineering teams.

Audience-Specific Update Formats

Leadership / Executive Updates

Leaders need: strategic context, progress mapped to goals, risks that require their intervention, and decisions awaiting their input.

Template:

Status: [Green / Yellow / Red]

Summary: [Single sentence — the one thing they must know]

Progress:
- [Outcome delivered, linked to OKR or strategic goal]
- [Milestone completed, with business impact noted]
- [Key metric shift]

Risks:
- [Risk statement]: [Mitigation underway]. [Specific ask if applicable].

Decisions required:
- [Decision topic]: [Options with a recommendation]. Needed by [date].

Upcoming milestones:
- [Milestone] — [Target date]

Guidance:

• Open with the conclusion, not the process. Leaders want "we launched X and it moved metric Y" rather than a list of meetings and tickets.
• Stay under 200 words. They will ask for more if they want it.
• The status color should reflect your honest assessment. Yellow signals healthy risk awareness, not failure.
• Only surface risks where you need help. Omit risks you are already resolving unless leadership needs visibility.
• Requests must be precise: "Approve vendor choice by Thursday" rather than "need alignment."

Engineering Team Updates

Engineers need: priority clarity, technical context, blocker resolution, and awareness of decisions affecting their work.

Template:

Shipped:
- [Feature or fix] — [Link to PR/ticket]. [Impact note if relevant].

In flight:
- [Work item] — [Owner]. [Expected delivery]. [Blockers if any].

Decisions:
- [Decided]: [Rationale]. [Link to ADR if available].
- [Pending]: [Background]. [Options]. [Recommended path].

Priority shifts:
- [What moved and the reason behind it]

On deck:
- [Upcoming items] — [Why these are sequenced next]

Guidance:

• Provide links to tickets, pull requests, and documents. Engineers want to drill into specifics.
• When priorities shift, explain the reasoning. Engineers invest more when they understand the why.
• Call out what is blocking them and describe your plan to clear it.
• Omit information that has no bearing on their work.

Cross-Functional Partner Updates

Partners across design, marketing, sales, and support need: advance notice of changes affecting their teams, preparation requirements, and channels for input.

Template:

Upcoming:
- [Feature or launch] — [Date]. [Implications for your team].

Action items for your team:
- [Specific request] — [Context]. Needed by [date].

Decisions finalized:
- [Decision] — [How it impacts your workflow].

Open for feedback:
- [Topic where input is welcome] — [How to submit it].

External / Customer-Facing Updates

Customers need: what is new, what is coming, how it benefits them, and how to take advantage of it.

Template:

What's new:
- [Capability] — [Benefit stated in customer language]. [Link or instructions].

On the roadmap:
- [Capability] — [Approximate timing]. [Why it matters to you].

Known issues:
- [Issue] — [Current status]. [Workaround if one exists].

Share your feedback:
- [Channel or method for submitting feature requests and feedback]

Guidance:

• Eliminate internal terminology. No ticket IDs. No architecture references.
• Frame everything around what the customer can now accomplish, not what your team built.
• Be candid about timelines without over-committing. "Later this quarter" is preferable to a date you might miss.
• Only include known issues if they are customer-visible and accompanied by a resolution path.

Traffic Light Status Framework

Definitions

Green (On Track):

• Work is proceeding according to plan
• No material risks or obstacles
• Commitments and deadlines will be met
• Reserve Green for when things are genuinely progressing well, not as a default

Yellow (At Risk):

• Progress has slowed or a risk has materialized
• Mitigation is in progress but the outcome remains uncertain
• Commitments may be missed without intervention or scope adjustment
• Flag Yellow early — the sooner risk is visible, the more options remain

Red (Off Track):

• Significantly behind schedule
• A major obstacle exists without a clear resolution
• Commitments will not be met without substantial intervention (scope reduction, additional resources, or timeline extension)
• Escalate to Red when your own options are exhausted and you need external help

Transition Rules

• Shift to Yellow at the first indication of risk, not once you are certain things have gone wrong
• Shift to Red when you have used every lever available to you and require escalation
• Return to Green only after the underlying risk is genuinely resolved, not merely paused
• Always document the reason when changing status: "Moved to Yellow because [explanation]"

Communicating Risk Effectively

ROAM Classification

• Resolved: the risk no longer applies. Record how it was eliminated.
• Owned: the risk is recognized and someone is actively managing it. Name the owner and the mitigation approach.
• Accepted: the risk is acknowledged but the team proceeds without active mitigation. Capture the rationale.
• Mitigated: actions have reduced the risk to a tolerable level. Document what was done.

Five-Step Risk Communication Pattern

1. Name it plainly: "There is a risk that [event] occurs because [cause]"
2. Size the impact: "If this materializes, the consequence is [specific impact]"
3. Estimate probability: "This is [likely / possible / unlikely] because [evidence]"
4. Describe the response: "We are addressing this by [actions taken or planned]"
5. State the need: "To further reduce this risk, we need [specific help or decision]"

Pitfalls to Avoid

• Burying risks inside positive updates. When a risk is significant, lead with it.
• Vagueness: "We might see some delays" — always specify what could be delayed, by how long, and why.
• Presenting risks without accompanying action plans. Every risk should include a mitigation or a request.
• Late disclosure. A risk surfaced early is a planning input. A risk surfaced late becomes a crisis.

Decision Records

Lightweight ADR Structure

Record significant decisions for institutional memory:

# [Decision Title]

## Status
[Proposed / Accepted / Deprecated / Superseded by ADR-XXX]

## Situation
What circumstances necessitate this decision? What constraints and pressures exist?

## Resolution
What was decided? State the choice clearly and unambiguously.

## Tradeoffs
What follows from this decision?
- Benefits gained
- Downsides or compromises accepted
- Future possibilities opened or closed

## Options Evaluated
What alternatives were considered?
For each: a brief description and the reason it was not chosen.

When to Create a Decision Record

• Strategic product choices (target market, platform support, pricing model)
• Major technical decisions (architecture direction, vendor selection, build vs. buy)
• Contested decisions where the team disagreed (preserve the reasoning for posterity)
• Decisions that narrow future flexibility (technology lock-in, partnership commitments)
• Decisions you anticipate will be questioned later (capture context while it is fresh)

Practical Tips

• Write the record close to the moment the decision is made, not weeks afterward
• Note who participated and who held final decision authority
• Document the surrounding context generously — future readers will lack today's situational awareness
• Records that prove wrong in hindsight are still valuable; add a "superseded by" reference
• Brevity is an asset. A single page is better than five.

Meeting Facilitation Patterns

Daily Sync / Stand-up

Goal: uncover blockers, synchronize efforts, sustain momentum.

Each participant shares:

• Work completed since the last sync
• Current focus
• Anything blocking progress

Running it well:

• Cap at 15 minutes. Park longer discussions for follow-up.
• Prioritize blockers — they are the highest-leverage item in any standup
• Track blockers explicitly and follow up until resolved
• Skip the meeting if there is nothing to synchronize. Protect people's time.

Iteration / Sprint Planning

Goal: agree on the work for the upcoming cycle and align on scope and priorities.

Agenda:

1. Retrospective glance: what shipped, what carried over, what was descoped
2. Priority declaration: the most important outcomes for the coming sprint
3. Capacity assessment: realistic bandwidth accounting for PTO, on-call duties, and meetings
4. Commitment selection: pull items from the backlog that fit capacity and priority
5. Dependency mapping: flag anything requiring cross-team or external coordination

Running it well:

• Arrive with a proposed priority ranking. Do not ask the team to sort from zero.
• Resist overcommitment. Delivering less reliably builds more trust than consistently missing targets.
• Every selected item needs a clear owner and unambiguous acceptance criteria.
• Surface items with hidden complexity or insufficient scoping.

Retrospective

Goal: reflect honestly on what worked, what did not, and what the team will change.

Flow:

1. Frame the conversation: recall the sprint goal and establish psychological safety
2. Collect observations: what went well, what fell short, what caused confusion
3. Identify patterns: find recurring themes and root causes
4. Select improvements: commit to 1-3 concrete changes for the next cycle
5. Close: acknowledge the team's candor

Running it well:

• Psychological safety is non-negotiable. People must feel safe to speak honestly.
• Direct attention toward systems and processes, never toward individuals.
• Cap action items at 1-3. More than that and nothing gets implemented.
• Revisit prior retrospective actions. If follow-through never happens, engagement erodes.
• Rotate the retrospective format periodically to maintain freshness.

Stakeholder Demo / Review

Goal: demonstrate progress, solicit feedback, reinforce alignment.

Flow:

1. Set context: remind stakeholders of the objective and what they saw previously
2. Live demo: show the working product, not slide decks
3. Data share: present any early metrics or user feedback
4. Structured feedback: guide questions toward specific topics
5. Forward look: preview what comes next and when the next review will occur

Running it well:

• Always demo the real product when possible. Slides are not substitutes.
• Direct feedback requests precisely: "What is your reaction to X?" outperforms "Any thoughts?"
• Capture feedback visibly and commit to addressing each item or explaining why it will not be addressed
• Clarify what type of feedback is actionable at the current stage