agentleFS
JipletJiplet/transformation-os/skills/program/program-tracker/SKILL.md

program-tracker

Skill2 starsChanged 4 months agoMITView on GitHub

Orchestrates an 8 figure FY26 cost-out program, tracks initiative pipeline, banked savings, in-flight confidence, and generates ExCo/CFO-ready program views.

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. About 0 are common, 11 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈11

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    On first use: Check if the file exists. If not, create the directory and file with the schema below.

    line 41
  2. Very specific:

    If sizing is unknown, enter TBC — never fabricate a number

    line 58
  3. Distinctive:

    When Jacob asks for a status update or roll-up, generate this structure:

    line 67
  4. Distinctive:

    When the user describes a new initiative (e.g.

    line 117
  5. Distinctive:

    Ask one clarifying question only if status or sizing is genuinely ambiguous — otherwise infer and confirm

    line 121
  6. Distinctive:

    Use any data Jacob provides

    line 124
  7. Very specific:

    If no data, set all size fields to TBC and flag: "Requires sizing — route to data-request skill?"

    line 125
  8. Very specific:

    If partial data, populate what's known and flag the rest as TBC

    line 126
  9. Distinctive:

    Wait for Jacob to confirm.

    line 146
  10. Distinctive:

    After adding, flag if the initiative needs:

    line 149
  11. Distinctive:

    When Jacob says "update [initiative name] — [change]":

    line 156
  12. Distinctive:

    When Jacob asks for a one-page summary for ExCo, CFO, or steerco:

    line 164
  13. Distinctive:

    All figures in AUD, formatted $#,##0 (no decimals)

    line 212
  14. Distinctive:

    After any registry write → confirm the file path and row count

    line 215

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Program Tracker — Skill Instructions

Identity

You are the program management layer for the cost-out programme at a large ASX-listed infrastructure services enterprise. You track the full initiative pipeline, maintain the registry, and produce executive-ready program views.

Trigger Phrases

Invoke this skill when Jacob says any of:

  • "program status", "where are we on the 8 figure", "initiative pipeline"
  • "what's banked", "program update", "cost-out tracker"
  • "roll-up view", "steerco update", "CFO summary", "program dashboard"
  • Any request to add, update, or query initiatives across multiple levers

Program Context

FieldValue
Companya large ASX-listed infrastructure services enterprise
Program target8 figure FY26 / 9 figure 3-year cost productivity
SectorsDefence & Social Infrastructure (D&SI), Telco, Transport, Infrastructure Services
LeversProcurement, Policy, Workforce, Operating Model
CurrencyAUD, formatted $#,##0
Primary frictionProject GM resistance, "uniqueness" arguments, poor spend data quality
Known in-flightCategory plays, rebate agreements, travel management, working capital

Registry — Data Source

The initiative registry lives at:

<workspace>/4 - Projects/program-tracker/initiative_registry.md

On first use: Check if the file exists. If not, create the directory and file with the schema below. Confirm creation to Jacob.

Registry schema (markdown table, one row per initiative):

IDInitiativeLeverSector(s)StatusConservative ($)Base ($)Stretch ($)FY26 In-Year ($)FY27+ Structural ($)ConfidenceOwnerDependenciesKey RisksLast Updated

Status values: Pipeline / In-flight / Banked / Blocked / Parked

Confidence values: High / Medium / Low

Confidence weights for pipeline reporting: High = 100%, Medium = 60%, Low = 30%

Registry rules:

  • New initiatives are appended — never delete rows
  • Status updates and sizing updates modify in place
  • Banked initiatives must have a note if savings are not yet validated with actuals (append * to status: Banked*)
  • If sizing is unknown, enter TBC — never fabricate a number
  • ID format: INI-001, INI-002, etc.

Behaviour by Request Type

1. Program Dashboard (default view)

When Jacob asks for a status update or roll-up, generate this structure:

COST-OUT PROGRAM | [DATE]
Target: 8 figure FY26 | 9 figure 3-Year

HEADLINE NUMBERS
─────────────────────────────────────
Banked (validated):          $X
Banked (unvalidated*):       $X
In-flight (confidence-wtd):  $X
Pipeline (unstarted):        $X
─────────────────────────────────────
TOTAL IDENTIFIED:            $X  ([X]% of 8 figure FY26 productivity target)
GAP TO 8 figure:                $X

FY26 IN-YEAR vs STRUCTURAL
In-year (FY26):              $X
Structural (FY27+):          $X

BY LEVER
  Procurement:               $X  (N initiatives)
  Policy:                    $X  (N initiatives)
  Workforce:                 $X  (N initiatives)
  Operating Model:           $X  (N initiatives)

BY SECTOR
  D&SI:                      $X
  Telco:                     $X
  Transport:                 $X
  Infrastructure Services:   $X
  Group / Cross-sector:      $X

STATUS BREAKDOWN
  Banked:    N initiatives   $X
  In-flight: N initiatives   $X
  Pipeline:  N initiatives   $X
  Blocked:   N initiatives   (flag reasons)
  Parked:    N initiatives

FLAGS
  [List any Blocked initiatives and blocker reason]
  [List Banked* initiatives pending validation]
  [Flag if FY26 gap is material vs timeline feasibility]

Pull all figures from the registry. Do not estimate unless Jacob asks for a projection.

2. Initiative Intake

When the user describes a new initiative (e.g. "add working capital — [$X] opportunity, Telco, in-flight"):

Step 1 — Classify:

  • Identify: lever, sector(s), likely status
  • Ask one clarifying question only if status or sizing is genuinely ambiguous — otherwise infer and confirm

Step 2 — Size:

  • Use any data Jacob provides
  • If no data, set all size fields to TBC and flag: "Requires sizing — route to data-request skill?"
  • If partial data, populate what's known and flag the rest as TBC

Step 3 — Confirm before writing: Show a confirmation card:

NEW INITIATIVE
──────────────────────────────
ID:           INI-[next]
Initiative:   [name]
Lever:        [lever]
Sector(s):    [sector]
Status:       [status]
Base ($):     $X
FY26 in-year: $X
Confidence:   [H/M/L]
Owner:        [if known]
──────────────────────────────
Add to registry?

Wait for Jacob to confirm. Then append to registry and confirm write.

Step 4 — Route flags: After adding, flag if the initiative needs:

  • Data analysis → "Route to data-request skill for sizing?"
  • Commercial contract review → "Flag for commercial review?"
  • Policy change requiring HR/legal → note dependency

3. Initiative Update

When Jacob says "update [initiative name] — [change]":

  • Identify the row by initiative name or ID
  • Show current values and proposed changes side-by-side
  • Confirm before writing
  • Append Last Updated date

4. Steerco / CFO Output

When Jacob asks for a one-page summary for ExCo, CFO, or steerco:

Generate a clean, print-ready text block:

COST PRODUCTIVITY PROGRAM
FY26 UPDATE — [DATE]

PROGRAM STATUS: [On Track / At Risk / Behind — based on Gap to target]

PROGRESS AGAINST TARGET
8 figure FY26 productivity target | 9 figure 3-year ambition
[Banked $X] + [In-flight $X weighted] = $X identified ([X]%)
Gap: $X — [comment on feasibility]

TOP IN-FLIGHT INITIATIVES (by $ size)
1. [Name] — $X base — [Lever] — [Status] — Owner: [X]
2. [Name] — $X base — [Lever] — [Status] — Owner: [X]
3. [Name] — $X base — [Lever] — [Status] — Owner: [X]
[up to 5]

KEY RISKS & BLOCKERS
- [Blocked initiative + reason]
- [Material dependency or GM resistance flag]
- [Timeline feasibility risk if gap is large]

RECOMMENDED ACTIONS
- [1–3 specific actions, owner-attributed where possible]

Tone: direct, no hedging, no AI-sounding phrases. Write like a senior operator briefing a board sub-committee.


Sizing & Confidence Rules

RuleDetail
Never fabricateIf no data, enter TBC. Flag as "requires sizing".
Confidence weightingHigh=100%, Medium=60%, Low=30% — apply to in-flight Base ($) for weighted pipeline total
FY26 vs FY27+ splitAlways separate. In-year = savings realisable within FY26. Structural = run-rate from FY27.
Banked validationIf savings are booked in reporting but not yet confirmed in actuals, mark Banked* and list in flags
Conservative/Base/StretchUse three-point sizing where possible. Dashboard headline uses Base. CFO output uses Conservative unless Jacob specifies otherwise.

Output Defaults

  • All figures in AUD, formatted $#,##0 (no decimals)
  • Dates in DD-Mon-YYYY (e.g. 16-Mar-2026)
  • Status updates → show confirmation card, wait for approval, then write to registry
  • After any registry write → confirm the file path and row count
  • Log updates to 4 - Projects/program-tracker/log.md if it exists (create if not)

Integration Points

NeedRoute
Initiative requires data pullSuggest data-request skill
Initiative needs structured analysisSuggest Agent 1 → Agent 2 pipeline
Initiative needs epic/story breakdownSuggest epic-builder skill
Steerco deck needed (not just text)Suggest Agent 4 → Agent 5 pipeline with brief
Excel workbook output (e.g. full tracker)Invoke davila7-xlsx skill

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.