succession_planning v1.0.0

name: "succession_planning" description: "Succession planning for critical roles: score role criticality, rate successor readiness honestly, build per successor development plans, name emergency cover, and audit bench strength with a runnable tool. Use when asked for a succession plan, critical role or key person risk review, leadership bench planning, successor readiness ratings, or an emergency or interim cover plan."

Succession Planning

Purpose

Keep critical work running when a key person leaves, is absent, or moves. This skill turns a list of names into a working system: scored critical roles, evidence based readiness ratings, development plans that close real gaps, a separate emergency cover sheet per Tier 1 role, and bench metrics a leadership team can review on a cadence. A list of names with no readiness evidence, no gaps, and no dates is not a succession plan.

Workflow

1. Build the critical role list from harm, not titles. Pull roles from org charts, incident history, audit duties, customer ownership, and single person dependencies. Ask what breaks if the role is vacant for 30, 60, and 90 days. Include specialist and individual contributor roles when the knowledge is concentrated, not only managers.

1. Score every candidate role with scripts/succession.py score or in bulk via CSV, using the rubric in references/scoring.md. Five factors on a 1 to 5 scale: impact, vacancy speed, knowledge concentration, relationships, and replacement difficulty. Tier 1 is 75 and above, Tier 2 is 55 to 74, Tier 3 is below 55. Bring evidence for each score to the review, not an opinion about the incumbent.

1. Write a success profile for each Tier 1 and Tier 2 role before naming anyone. Outcomes the role must deliver in the first 90 days, at six months, and at one year, plus the technical, people, and stakeholder skills that delivery needs. Readiness is judged against this profile, never against the current holder's personality and never against success in a smaller role alone.

1. Name successors per role and rate readiness using the ratings in references/scoring.md: ready_now, ready_6_12mo, ready_1_2yr, ready_3_plus_yr, emergency_only, or none. Aim for more than one successor per Tier 1 role at differing readiness. For a class of similar roles, build a talent pool instead of a fragile one to one chart. Record flight risk per successor honestly.

1. Calibrate in a group session using the agenda in references/development-and-emergency.md. Separate performance from potential when a 9-box discussion helps. Require work at target scope as evidence, downgrade a rating two raters cannot support, and record dissent. Roles rated none and Tier 1 roles with no ready_now successor are findings: they get a hiring readiness step, a restructuring option, or an accepted risk note signed by the decision maker.

1. Write one development plan per successor per role, following references/development-and-emergency.md. Name the two or three gaps that block readiness, then assign work in the 70-20-10 pattern: about 70 percent stretch assignments, rotations, or acting stints with measurable outcomes, 20 percent mentoring and coaching with a named person and cadence, 10 percent formal learning tied to a named gap. Add proof of readiness and a review date. Quarterly reviews for Tier 1 successors, twice a year for Tier 2.

1. Fill an emergency cover sheet for every Tier 1 role, separate from the long term plan. Standing acting appointee plus first and second backups, who notifies whom, day one authority and its limits, the first 72 hours checklist, and where the operations manual and contact lists live, with no passwords or secrets in the sheet. The emergency pick keeps operations running for weeks. The long term pick is chosen for years. They may be different people.

1. Audit the bench with the bundled tool and report the metrics.

bash
   python3 scripts/succession.py score --role "Plant Operations Lead" --impact 5 --vacancy_speed 5 --knowledge 4 --relationships 3 --replacement_difficulty 4
   python3 scripts/succession.py report --roles references/example-roles.csv --successors references/example-successors.csv
   

The report exits with code 1 when a critical flag exists, so it can gate a review. Report coverage ratio, Tier 1 ready now percentage, bench points per role, and every critical flag. Worked example files are in references/example-roles.csv and references/example-successors.csv: the Plant Operations Lead scores 87.0 Tier 1 and flags critical because its bench has development and emergency cover but nobody ready now.

1. Set the review cadence and triggers. Full review at least twice a year for Tier 1 and Tier 2 roles, annually for Tier 3 pools, and immediately after an exit, a restructure, a major strategy change, or a successor leaving. Ratings older than 12 months are stale and get re-rated. Close the loop by checking whether last cycle's development assignments actually happened, and re-rate down when they did not.

Output Contract

Deliver a succession packet containing, at minimum:

• The scored critical role register: role, five factor scores, criticality score, tier, incumbent vacancy risk note, and the evidence behind the scores.
• A successor table per Tier 1 and Tier 2 role: candidate, readiness rating, flight risk, named gaps, and evidence for the rating.
• One development plan per successor, with assignments, owners, dates, proof of readiness, and review date.
• An emergency cover sheet per Tier 1 role, with acting appointee, two backups, authority limits, and the first 72 hours checklist.
• A bench audit: coverage ratio, Tier 1 ready now percentage, critical flags, and the agreed actions for each flag, with owners and dates.
• The next review date and the triggers that force an early review.

Operating Rules

• Score the role, not the person. A strong incumbent can hide a critical role. A weak incumbent does not make a low impact role critical.
• Honest ratings only. Inflated ready now labels are the main way succession plans fail in private while looking fine on paper. When in doubt, rate one level lower and name the evidence that would raise it.
• Never treat succession names as promises. Being listed is not a guarantee of promotion, and the plan should say so in plain language wherever it is shared. Handle the file as confidential HR material: share on a need to know basis, never paste secrets, passwords, or private personnel notes into templates or tools.
• Keep emergency cover and long term succession separate, and keep both current. An emergency sheet nobody can find at night is the same as no sheet.
• Development must be real work, not coursework alone. If a plan has no stretch assignment, no acting time, and no delegated authority, it will not close a readiness gap.
• Count a pool fairly. Look at who never gets nominated and ask whether the evidence rule, not the person, is the barrier. Widen evidence sources (work samples, project outcomes, structured input from more than one manager) before widening exceptions.
• Cut unverified numbers. Do not quote bench, turnover, or cost of vacancy statistics you cannot trace to a source the user can inspect. Estimate cost of vacancy only from inputs the user supplies, and label it an estimate.