Releases: systima-ai/project-delivery-framework
v0.1.4 — transcript cleaning, call debrief, engagement-state tracking
Install: npx @systima/project-delivery-framework
This release rolls up 0.1.3 (previously committed but never published) and 0.1.4.
0.1.4 — Added
pdf-clean-transcript— raw machine transcript (Teams / Zoom / Otter / Whisper / single-channel recorder export) to cleaned, speaker-attributed markdown with normalised terms and explicit[?]uncertainty markers. Standard "Editor's notes" preamble makes every inference visible. Output is source material (_transcripts/), not a canonical artifact.pdf-debrief-call— after-action review of a consequential call (discovery / stakeholder / negotiation / interview). Evidenced dimension ratings, intelligence capture, owned follow-ups. Honest by design (mandatory weakness).- Canonical-facts register (
00-constitution/FACTS.md) — single source of truth for the client/org name, key people, and a glossary of normalised terms; confirmed vs[?]unconfirmed (ARCHITECTURE §8.2). - Engagement lifecycle status —
lifecycle:(prospective | go | active | on-hold | closed) in the engagement README;pdf-helpgates its recommendations on it (ARCHITECTURE §8.3). pdf-help stalemode — advisory freshness sweep: charter-revision drift, source-newer-than-artifact, unresolved[TBC]/[?]markers.
0.1.4 — Changed
- Stub-aware
pdf-help— newdone_whenCSV column distinguishes an authored artifact from a freshly-scaffolded stub (e.g. charter is "done" only atcurrent_revision>=1). Backward-compatible. [?]convention formalised alongside[TBC](ARCHITECTURE §8.1);pdf-elicitwalks both.
0.1.3 — Added
pdf-create-measurement-plan(Petra / Planning) — GQM measurement baseline tracing metrics to charter success criteria, with a fatal-on-fail coverage check.
v0.1.2 — Multi-tool support (Claude Code + OpenCode)
What's new
PDF now works natively in Claude Code, OpenCode, and any tool following the .agents/skills/ convention.
Installer changes
The installer now writes skills to three locations:
.claude/skills/pdf-*/— primary; discovered by Claude Code and OpenCode.agents/skills/pdf-*/— tool-agnostic mirror.opencode/commands/pdf-*.md— OpenCode slash-command wrappers so PDF commands appear as/pdf-help,/pdf-agent-mobilizeretc. in OpenCode's TUI
Install
npx @systima/project-delivery-frameworkAfter install, in your AI tool of choice:
/pdf-help # see what to do next
/pdf-engagement-init # scaffold a new engagement
Or talk to an agent: hey Marcus, hey Ronan, hey Helena, etc.
Other
scripts/generate-opencode-commands.js— build script that generates the OpenCode wrappers from each skill's frontmatter description.- README and install-flow output updated to reflect the three discovery locations.
opencodeadded to package keywords.
See CHANGELOG.md for full details.
v0.1.1 — scope rename to @systima
What changed
Package renamed to @systima/project-delivery-framework for publishing under the Systima npm org.
New install command
npx @systima/project-delivery-frameworkUpdates
package.jsonname:project-delivery-framework→@systima/project-delivery-frameworkREADME.mdandbin/cli.jsupdated throughout (12 occurrences)- README badges added: npm version, GitHub release
- The unscoped package at
npmjs.com/package/project-delivery-frameworkhas been deleted; only the scoped version exists going forward
Skill set
Unchanged from v0.1.0 — same 62 skills (10 agents + 52 workflows / utilities). This release is a package-rename only.
See CHANGELOG.md for full v0.1.0 details.
v0.1.0 — Initial public release
Project Delivery Framework v0.1.0
Initial public release. Spec-driven, AI-assisted operating system for delivery and engagement managers — the BMAD sibling for the delivery-management layer.
Install
npx project-delivery-frameworkInstalls into the current directory. After install, open in Claude Code and run /pdf-help or talk to an agent ("hey Marcus", "hey Ronan", etc.).
What's in v0.1.0
10 stage agents covering the full engagement lifecycle
| Stage | Agent | Scope |
|---|---|---|
| 01 — Shaping | Sofia | Qualification (MEDDIC), opportunity shape, ROM estimate (cone of uncertainty), SoW draft |
| 02 — Mobilization | Marcus | Charter, RACI, governance, comms plan, kickoff deck |
| 03 — Planning | Petra | Plan (Mermaid Gantt + phase table), capacity (matrix + rows), budget baseline (monthly + milestone), estimate challenge |
| 04 — Execution | Ronan | Weekly status (Teams/Slack-paste-friendly), RAID, standup digests, velocity, blocker triage |
| 05 — Governance | Helena | Steering pack (MARP), exec summary (1-page strict), stakeholder updates (9 archetypes), escalation memos (SCQA) |
| 06 — Risk & Change | Klaus | Risk deep-dives, change requests (granular taxonomy with classic inference), mitigation, escalation decisions |
| 07 — Quality | Quinn | SDLC, QA, SysEng, Secure SDLC health cards (RAG-rated, compliance cross-check) |
| 08 — Commercial | Theo | Budget tracker (weekly + monthly), margin analysis (period + milestone), change orders, commercial-model review, invoice backup |
| 09 — People | Iris | Team health, attrition risk, ramp plans, 1:1 prep, performance conversations (Radical Candor) — confidential by default |
| 10 — Closure | Felix | Closure checklist, retrospective (internal + client-shareable), lessons (propagated to practice library), case study, handover |
Plus
- 5 cross-cutting utilities:
pdf-red-team,pdf-translate,pdf-decision-log,pdf-audit-log,pdf-elicit - 52-row CSV skill index driving
pdf-help's next-action recommendations - Living
CHARTER.mdas the engagement constitution; every artifact reconciles to its current revision - Audit-by-default: per-run JSONL records of prompt + model + output hash
- Adversarial red-team gate on every outward-facing artifact
- 62 skill folders total
Design influences
- BMAD-Method — skill anatomy, agent/workflow split, CSV index pattern
- Spec Kit — the constitution pattern
- Anthropic Knowledge Work Plugins — file-based, locally-runnable philosophy
- ArcKit — audit-ready artifact structure
Verify
npx project-delivery-framework statusQuick start
npx project-delivery-framework # install into current directory
# then in Claude Code:
/pdf-engagement-init # scaffold a new engagement
/pdf-help # see next required action
# or: "hey Marcus", "hey Ronan", etc.Documentation
ARCHITECTURE.md— full design spec (20 sections, decisions log)README.md— what / who / howCHANGELOG.md— per-stage build history
Known scope for v0.2+
- API sync providers (Jira, ADO, GitHub, Confluence, Calendar, Slack/Teams, time-tracking) — architecture in
ARCHITECTURE.md§20 - Cross-engagement portfolio view
- Practice-library search during engagement init
Licence
MIT