Skip to content

CCPM Project Data

The CCPM Skill attempts to create CCPM project schedules.

Files formats

tasks.csv

tasks.csv — id, name, duration_safe, duration_aggressive (optional), predecessor_ids, resource_ids, url (optional)

id,name,duration_safe,predecessor_ids,resource_ids,url
A,Spec,10,,blue,https://example.com/wiki/spec
B,Build,20,A,green,https://example.com/tickets/build
F,Commission,10,D;E,red,
  • predecessor_ids: semicolon-separated links. A bare id is Finish-to-Start; typed links with lag are supported: A:SS+2, A:FF, A:SF. The network must be acyclic; multiple entry or exit tasks are fine (the scheduler anchors them to synthetic Start/Finish milestones).
  • Every task needs a positive duration and at least one resource — a task without a resource cannot contend for capacity, so it cannot participate in critical chain identification.
  • If duration_aggressive is missing, the skill applies the classic 50% cut to duration_safe. If your estimates are already aggressive, say so in the prompt — it changes buffer sizes by 2x.
  • url: optional link to a ticket/wiki page; it is carried into the outputs.
  • tags, colour (optional, written by our-planner's Network → Export CCPM Network...): comma-separated task tags and the task's canvas colour. The scheduler currently ignores unknown columns (passthrough to schedule.csv is planned); our-planner's import reads them back if present.

resources.csv

resources.csv — id, name, capacity, url (optional), emails (optional); capacity defaults to 1.

A list of the project resources and default capacity.

  • url: optional link to the resource's profile/ticket queue/etc.
  • emails: optional contact address(es) for the resource, used by reports/exports to reach them directly - scheduling itself ignores this column entirely. One or more addresses, separated by , or ;. If you're hand-editing this file (outside our-planner's own export), prefer ; between multiple addresses: a bare , only survives round-tripping through a CSV reader if the whole cell is quoted (e.g. "alice@example.com,bob@example.com"), and a text editor or careless spreadsheet edit can easily leave that comma unquoted - which silently shifts every later column on the row rather than failing loudly. our-planner's own exports always quote correctly either way.

calendar.csv

calendar.csv (optional) — availability overrides:

resource_id,from,to,capacity
green,2,4,0
red,0,10,0

Overrides a resource's capacity on the day range [from, to). The bracket notation is deliberate — it is the mathematical convention for a half-open interval: the square bracket [ means from is included, the round bracket ) means to is excluded. So green,2,4,0 means green is unavailable on days 2 and 3, and back at work on day 4 — the range covers to − from days, never to itself.

Why half-open? It matches how the schedule itself works: a task with start=2, finish=4 occupies days 2 and 3 too, so a calendar row and a task span with the same numbers cover exactly the same days, and adjacent ranges like [0,5) and [5,10) butt together without overlapping or leaving a gap.

Days are working-day offsets from day 0 (the same axis as the Gantt chart). capacity 0 = unavailable (holiday, another project); a higher value models temporary extra capacity (e.g. a contractor). Outside the listed ranges, the resource's default capacity from resources.csv applies. Ranges for the same resource must not overlap.

our-planner's own Network → Export CCPM Network... and Network → Schedule with CCPM... populate the "another project" case automatically: resources are shared globally across every project in a plan file, so a resource's exported capacity is reduced by its load from every other project's tasks over the same days (see User Guide → Scheduling with CCPM) — hand-authoring these rows yourself is only needed for leave/contractor windows a project's own tasks can't already account for.

Tasks run contiguously — they never pause across an outage, so a task is scheduled entirely before or entirely after it.

schedule.csv

Every task and buffer with start/finish day offsets, chain membership (critical, feeding-n), and link notation. Buffers attach via CCPM-specific :PB/:FB link types (they behave differently from work during execution — slippage consumes them rather than pushing them).

our-planner's Network → Import CCPM Schedule... additionally reads optional tags and colour columns (color accepted as an alias) if present, and tags every imported row ccpm.

summary.md

— critical chain sequence, project duration, buffer sizes, and the promised completion date (= end of the project buffer). Task/resource urls become clickable links here.

gantt.png

The chart: critical chain in cross-hatched dark red, feeding chains colored, buffers hatched gold/khaki with a commitment-date diamond, dependency arrows (non-FS links labeled), and a resource-utilization panel where red means over capacity and grey hatching means unavailable.