DmitryDF
@DmitryDF
Public Skills
close
by DmitryDF
Close session - create diary entry, archive logs, git commit
q-update
by DmitryDF
Update an installed Q framework to a newer release from a local clone of the Q repo — shows what changed, applies only the delta, and never overwrites the files that belong to you. Use when the operator asks to update Q, pull a newer Q release, or check whether their Q install is current. Not for the first install; that is INSTALL.md.
ninja-fix
by DmitryDF
Ultra-fast-track entry into the 5-phase workflow. Resolves a concrete upstream-surfaced code error end-to-end in one session via a 5-step locked contract (Diagnosis → Guiding Policy → Coherent Action → independent validator → user-confirmation hard-gate). No spine, no plan file, no handoff. Replaces legacy /quickfix.
execute-plan
by DmitryDF
Plan-execution orchestrator (v2) — walks a locked _PLAN slice register across session boundaries via a code-derived typed cross-session handoff (continue-the-run / plan-this-slice / implement-this-slice) carried on active-run.json's pending_handoff and routed through the live execplan-session-ack gate. The v2 cross-session dispatch loop is the single dispatch entry point; the hands-off arm reuses the in-session spawn + two-layer-verify + commit path. Main-session use only.
work-done
by DmitryDF
Mark a ship event done — composes the four-surface atomic write (TODO line, spine slice-register, spine Sessions log, conditional retire-marker) through L's verify_write per surface, with crash-safe resume.
work-start
by DmitryDF
Take a TODO or _thought into in-progress — acquires the per-topic lock, starts the appropriate phase, and synchronizes status across TODO.md / topic-state JSON / spine slice-register.
plan-followups-review
by DmitryDF
Walk a set of already-verified observations one at a time into triaged TODOs. For each observation the operator sees a plain-words "why this matters", an optional code-first /recommend with trade-offs, then an explicit accept/edit/discard. A follow-up TODO is created only on explicit accept and only through the framing gate — never silently, never in a batch. Reusable primitive; the caller hands pre-verified observations and reads back a disposition summary. Main-session use only.
plan
by DmitryDF
AI orchestrator — coordinates Phase 3 (planning). Consumes the upstream chain from a topic _thought (Mode A full upstream / Mode B chain present + invoke /solution-design first / Mode C Ninja-Plan no spine) BEFORE EnterPlanMode fires. Sequences per-transition coherency dispatches (factcheck-plan-step 0A/0B2/0C) and a final coherency check (factcheck-plan-coherency) at each structural transition with explicit per-spawn model pinning (Opus = primary thinker, Sonnet = support; early-exit on Round-1 PASS). After plan approval (ExitPlanMode), hard-gates the next move via an AskUserQuestion (a) handoff-prompt / (b) /close+/clear / (c) continue-in-session triple driven by the post-plan-uxgate.sh + check-post-plan-pending.sh marker pair.
solution-design
by DmitryDF
AI orchestrator — coordinates Phase 2 (solution design + delivery slicing). Dispatches N parallel solution-designer (B1) calls with varied parameters, ranks them via an Opus selector, post-verifies the ranking via /double-check (class check), then auto-selects the verified top-ranked candidate and elaborates it to full requirement coverage (coverage-gated via /double-check (class check) until 0 gaps), presents the coverage-complete design in the topic's <base>_DESIGN.md (status pending-approval) for the operator to approve, invokes solution-slicer (B2) on approval, final-verifies via /double-check (class gate), and flips the Design to chosen. On ESCALATE at any post-verify seam, surfaces a two-option choice — (a) one more round, (b) proceed as is. Stateless skills, orchestrator-owns-persistence (C5).
solution-slicer
by DmitryDF
Rule-driven delivery-slicing assessor + multi-perspective recommender for
solution-designer
by DmitryDF
AI adapter — designs ONE candidate Agreed Solution Description (architecture + UX folded) from a clarified # Discovery section. Called N-parallel by /solution-design orchestrator; also invokable by /plan Slice I fallback or standalone CLI. Stateless — never writes _thought.
prompt-for-handoff
by DmitryDF
Generates a PE-conformant session-continuity prompt on demand, at any point in any session — output to chat, and persisted to the target file's ## Next Session Prompt section when a file resolves. Use when the operator asks for a handoff prompt, a continuation prompt, or a next-session prompt, or says "hand this off", "write the handoff", or "prep the next session". Also called internally by /clarification Step 10, /close section 1d, and /work-done's M7 last step. Not for writing the diary or closing a session — that is /close.
clarification
by DmitryDF
Clarification phase — 10 hard-gated steps (+ Step 1b problem framing) that turn a raw idea into a 4-section _thought file with 4 locked fields in # Discovery, ready for /solution-design. Opens with three confirmation moments — the idea reflected back as what was understood (Step 1), the problem framed from that confirmed idea (Step 1b), then the full reflection drawn from both (Step 2) — so a misread is caught where it is cheapest to fix. Supports resume mode against an existing _THOUGHT.md.
research
by DmitryDF
Runs the one research methodology — scope framing and approval, a registered pipeline manifest, evidence-based findings written to a _RESEARCH.md as they land, fact-check, and a claims register — whoever calls it. Use when the operator types /research, or when /clarification Step 6 or /work-decode reaches its research step and calls this skill with a predefined scope. Fires only on an explicit call; phrasings like "what do you think" or "pros and cons" get an OFFER to run it, never a run. Main-session use only (it holds the shell and asks the operator); the shell-less research agent is its adapter, never its entry point.
opportunity-tree
by DmitryDF
Opportunity Solution Tree (OST) sub-skill — given a Problem statement, surface a candidate Desired Outcome by decomposing the problem space into opportunities and tracing each back to the operational outcome it implies. KL-primary via Cagan (Inspired ch.7) + Patton (User Story Mapping ch.13); URL-grounded via Teresa Torres OST reference. Called by /clarification Step 1b via opt-in --via-opportunity-tree flag.
lean-analytics-metrics
by DmitryDF
Validate candidate OMTM + secondary metrics against Lean Analytics' 4 good-metric properties + actionability. Called from /clarification step 8 and any other workflow that needs to gate metric choice before committing.
outcome-framing
by DmitryDF
Cagan-grounded outcome framing — turn an own-words solution + diagnosis into a Cagan-framed Outcome statement (output → outcome shift). Called from /clarification step 7 and any other workflow that needs to draft a desired-outcome statement from a problem + a proposed solution.
session-start
by DmitryDF
Start session - git check, read diary, check TODOs, check reminders
challenge
by DmitryDF
Stress-test an artifact against AI agree-bias. Returns critiques or open questions plus a hard-gate handoff. Main-session use only.
recommend
by DmitryDF
Recommend a single bounded forward move with trade-offs and concerns. Main-session use only.
double-check
by DmitryDF
Independent factcheck of any AI recommendation. Invokes ≥1 checker(s)
work-frame-and-create-todo
by DmitryDF
Frame a piece of work using the 4-component convention (Problem / Context / Guiding policy / Master plan) and add it to the project's TODO.md through the framing validator. Use when a user wants to create a [Thought]-tagged TODO item or any item that must pass the framing gate.