Milestones — Index
Task targets captured from working sessions. What was agreed, what has to get built, and which part of the map each target lands on.
Everything on this page is Planned — these are commitments from meetings, not shipped surface. The roadmap map shows what exists today.
| Milestone | Session | Targets | Status |
|---|---|---|---|
| Content Engine & Content Studio Requirements | 22 August 2026 | 19 targets | ⬜ Planned |
| Social Discovery, Factsheet Analysis & Video Generation | Social media discovery, financial data, creative/video assets | 7 decisions, 8 targets | ⬜ Planned |
Targets — the convention
The ## Targets checklist in each milestone doc is the schedulable backlog, and
the only thing the weekly charter draws from:
## Targets
- [ ] Apply blog tonality filters before generation, not after. <!-- id:CS-01 team:1 -->
The trailing <!-- id: --> is optional and invisible in the rendered page. It
gives a target a stable identity so a reworded line does not read as a new one;
without it the id is derived from the text, and rewording drops any selection
held against it. An optional team:1 in the same comment assigns the target to
Team 1's column on the Charter page; untagged targets belong to Team 2. An
optional dev:N after it splits the team's card by developer.
Two rules keep the list honest:
- Milestone docs only. Workflow specs describe their gaps and blockers in
prose under
## Gapsand## Blocked on, and that is where those stay. Most of what they name waits on integrations nobody can schedule — CRM, a channel provider, a listening feed — so promoting them to checkboxes would fill the charter with lines that cannot be worked on. - Targets carry no owner, estimate or due date. Everything the exported charter prints is read from the doc. Progress is the one column it does not own: it ships as a dropdown for the team to fill in.
The Charter page lists every open target, remembers what is ticked,
and exports the week's selection as .xlsx.
How the two relate
They are the two halves of the same sentence. Milestone 2 widens the front of the pipeline (new sources: social discovery, PDFs, factsheets) and extends its tail (script to video asset to published post). Milestone 1 tightens what happens in the middle — how a piece of content is configured, generated, scored and reviewed.
MILESTONE 2 MILESTONE 1 MILESTONE 2
(widen the front) (tighten the middle) (extend the tail)
──────────────── ──────────────────── ─────────────────
social discovery
PDF / factsheet ──────▶ brand guidelines
extraction content buckets
| dynamic inputs
insight extraction ──────▶ generation
|
AI / quality scoring
SEO optimisation
review
|
└──────────────▶ video asset generation
video editing
social media management
Separation that both sessions agreed on
Analysis and generation stay separate solutions joined by a data interface, not one workflow. Factsheet analysis emits insights; the Content Studio consumes them. Either side can be rebuilt without touching the other.
Where these land on the existing map
| Target area | Nearest built thing | Milestone |
|---|---|---|
| Social media discovery | Discovery Agent — news only today | 2 |
| PDF / factsheet extraction | Nothing. New capability | 2 |
| Script to video | Creative Agent stops at the script | 2 |
| Blog tonality, keywords, scoring | Creative Agent blog path | 1 |
| Brand + product guidelines | Intelligence layer — tenant-level only today | 1 |
| Content buckets, carousels, charts | Creative Factory | 1 |