project-management-raid-log

RAID Log

A RAID log is the single canonical place for the things that could derail a project. The four columns of the acronym aren't the same kind of thing — distinguishing them is the work.

• Risk — might happen. Probability × impact. Owned by someone who is mitigating.
• Assumption — currently believed true. If it turns out false, the plan changes. Owned by someone who is validating.
• Issue — is happening. Already real. Owned by someone who is resolving.
• Dependency — something the team needs from another team / vendor / process to make progress. Owned by someone who is chasing it.

The skill produces a structured xlsx workbook (because most stakeholders open xlsx, and filtering/sorting matters), and explains how to maintain it across the project lifecycle.

When to use this skill

• New project kickoff — establish the log.
• Mid-project — add or update entries as new things surface.
• Audit / steerco prep — review aging entries, close stale ones, surface tops to status.
• Conversion — turning a Slack thread or meeting note into proper RAID entries.
• Roll-up — summarizing for pm-weekly-status.

For the actual xlsx generation, lean on the xlsx skill — this skill defines the schema and rules; the xlsx skill handles file production.

Workbook structure

One workbook. Four sheets, one per RAID type. Plus a Summary sheet that aggregates open items by severity.

RAID-Log-{ProjectName}.xlsx
├── Summary       (read-only roll-up, formulas only)
├── Risks
├── Assumptions
├── Issues
└── Dependencies

Risks sheet columns

Column	Notes
ID	R-001, sequential. Never reuse.
Date raised	yyyy-mm-dd
Title	Short, scannable. "Contentful migration may slip Phase 2 launch."
Description	1-3 sentences. What might happen and why.
Probability	1 (rare) – 5 (almost certain)
Impact	1 (negligible) – 5 (severe)
Severity	Probability × Impact (formula). Color-banded: 1-6 green, 7-14 yellow, 15-25 red.
Mitigation	What we're doing to reduce probability or impact.
Contingency	What we'll do if it happens.
Owner	One person. Not "the team."
Status	Open / Mitigating / Realized (became an Issue) / Closed (mitigated or expired)
Last updated	yyyy-mm-dd
Notes	Free text. Date-stamped entries: 2026-04-15: ...

Assumptions sheet columns

Column	Notes
ID	A-001
Date raised	yyyy-mm-dd
Assumption	The thing we're treating as true. "Client will provide content for 12 product explainers by 2026-06-01."
Why it matters	What plan element depends on it.
Validation method	How we confirm. "Confirmed in writing by client content lead."
Validation deadline	yyyy-mm-dd. By when we need to know.
Owner	One person.
Status	Unvalidated / Validated / Invalidated (became a Risk or Issue) / Closed
Last updated	yyyy-mm-dd
Notes	Date-stamped entries.

Issues sheet columns

Column	Notes
ID	I-001
Date raised	yyyy-mm-dd
Title	"Contentful API rate limit hit during initial bulk import."
Description	What is happening and what it's blocking.
Impact	1-5
Urgency	1-5
Severity	Impact × Urgency (formula). Color-banded.
Resolution plan	The steps being taken.
Owner	One person.
Target resolution	yyyy-mm-dd
Status	Open / In progress / Blocked / Resolved
Last updated	yyyy-mm-dd
Notes	Date-stamped entries.

Dependencies sheet columns

Column	Notes
ID	D-001
Date raised	yyyy-mm-dd
Dependency	"Client marketing approval on disclaimer wording."
Type	Internal Slalom / Client / Third-party vendor / Regulatory
Needed by	yyyy-mm-dd
Why	What it unblocks.
Owner (chasing)	One person on the project team.
Counterparty	Who owns it on the other side.
Status	Identified / Requested / In progress / Received / Blocked
Last updated	yyyy-mm-dd
Notes	Date-stamped entries.

Summary sheet (read-only)

Pulls counts and tops from the four sheets:

• Open Risks by severity band (count of red / yellow / green).
• Top 5 risks by severity (live formula).
• Open Issues by severity band.
• Top 5 issues by severity.
• Unvalidated Assumptions whose validation deadline is within 14 days.
• Dependencies overdue (Needed by < today, Status ≠ Received).

This sheet is what gets screen-shared in steercos. Keep it free of formatting clutter.

Severity scoring conventions

Use 1-5 scales to keep math simple. A 5-point scale gives meaningful granularity without false precision.

Probability (Risks):

1. Rare — surprising if it happened.
2. Unlikely — could happen, no current signal.
3. Possible — coin flip / depends on factors not in our control.
4. Likely — current signals suggest it will.
5. Almost certain — would be surprising if it didn't.

Impact (Risks & Issues):

1. Negligible — no schedule, scope, or budget effect.
2. Minor — hours of rework, no stakeholder visibility.
3. Moderate — days of rework or one stakeholder ask.
4. Major — sprint-level slip or material scope cut.
5. Severe — phase or release at risk; executive escalation.

Urgency (Issues):

1. Whenever — no time pressure.
2. This sprint — should be addressed in current sprint.
3. This week — within 5 working days.
4. Today — within 24 hours.
5. Now — work blocked until resolved.

The math: severity = factor1 × factor2. Range 1-25. Color bands:

• 1-6: green
• 7-14: yellow
• 15-25: red

Stakeholders learn this scale week over week — don't change it mid-project.

Maintenance rhythm

Daily: anyone can add an entry. Use Slack channel or PR template that says "raised in RAID as R-NNN."

Weekly (Thursday before status):

• Owner reviews their entries.
• Update Last updated and Notes for any movement.
• Close anything resolved or expired.
• Surface top severity items into pm-weekly-status.

Monthly / mid-phase:

• Audit aged Open items. Anything Open >30 days without a Notes update is a smell — chase the owner or close.
• Re-score items where the situation has changed.
• Move realized Risks to Issues; move invalidated Assumptions to Risks or Issues.

Lifecycle transitions

The four types are not silos. Things move:

• Risk → Issue when it materializes. Add a note on the Risk closure (Realized — see I-NNN). Open a fresh Issue with a backlink.
• Assumption → Risk when it's looking shaky. Same backlink pattern.
• Assumption → Issue when it's already been falsified.
• Dependency → Issue when "Needed by" passes and it's still not delivered.

Backlinks are how the log stays auditable.

Common failure modes

• One spreadsheet with mixed types. Risks and issues are different — don't merge them.
• "The team" as owner. Always name a person.
• Severity stuck at 3/3 forever because nobody re-scores. Re-score when status changes.
• Issues that should be Risks (or vice versa). If it hasn't happened yet, it's a Risk.
• Closed items deleted from the sheet. Don't — keep the audit trail. Filter by Status instead.
• 100 entries, all "Open." Most of these are stale. Close aggressively; reopen if needed.
• "Mitigation: monitoring." That's not mitigation, that's denial. Push for a real action.

Producing the xlsx

When asked to generate the workbook, use the xlsx skill to create the file with:

• Headers frozen on row 1
• Data validation on Status, Type, Probability, Impact, Urgency columns (dropdowns)
• Conditional formatting on Severity columns (color bands)
• Summary sheet formulas as described
• Save as RAID-Log-{ProjectName}.xlsx (or project-appropriate name) in the workspace folder

For ongoing maintenance via this skill, accept input as either:

• A markdown table (paste from Slack, doc, meeting notes) → produce updated xlsx.
• An updated xlsx → re-summarize and surface changes for pm-weekly-status.

How this skill relates to others

• Feeds directly into pm-weekly-status (Risks & Issues table, Decisions Needed).
• Items often spawn user stories — see pm-user-story.
• File production handled by the xlsx skill.
• For requirements gathering that surfaces assumptions, see forthcoming pm-requirements-discovery skill.

Source frameworks

• Josephs & Rubenstein, Risk Up Front — primary source for the R column. RUF meeting structure, familiarity scoring, cause→effect→impact format, RAP-sheet structure, the chicken test, the <80% escalation rule. See ../../references/book-risk-up-front.md.
• Burtonshaw-Gunn, Essential Tools for Management Consulting — 4Ts Risk Response (Treat, Transfer, Terminate, Tolerate); standard risk-management process. See /80_Skills_and_Agents/consulting/references/book-essential-tools-burtonshaw.md.

Templates this skill uses

• ../../references/template-risk-up-front.md — primary deliverable for kickoff RUF