All authors
Claude Skills by Allan-Nava
github.com/Allan-Nava6 skills0 installs1 views
- HandoffHandoff — the note a fresh session with zero context reads to continue this work. Load it first, before any other tool call, whenever the ask is to summarise or dump what we found or the state of a session, an investigation, a refactor or a feature into a file — HANDOFF.md, a note before /clear, a handover doc for a teammate — to compress research for a new agent, to brief a subagent or another agent on a task, or to rewrite a plan step so a stranger can execute it without questions. The mate...Votes: 0GitHub stars: 2
- Qrspi NewCodex CLI: bootstrap a QRSPI task — create thoughts/<task-id>/ from the phase templates and run the Questions phase. Use only when the user invokes it by name — `$qrspi:qrspi-new` — never because a prompt resembles it; in Claude Code the same command is /qrspi:new.Votes: 0GitHub stars: 2
- Qrspi NextCodex CLI: detect which QRSPI phase a task is in and emit the exact prompt for the next one. Use only when the user invokes it by name — `$qrspi:qrspi-next` — never because a prompt resembles it; in Claude Code the same command is /qrspi:next.Votes: 0GitHub stars: 2
- Qrspi ReviewCodex CLI: review one QRSPI artifact for judgement-level weaknesses — finished-but-wrong, not unfinished — and report; edits nothing. Use only when the user invokes it by name — `$qrspi:qrspi-review` — never because a prompt resembles it; in Claude Code the same command is /qrspi:review.Votes: 0GitHub stars: 2
- QrspiRun a coding task through the QRSPI workflow — Questions, Research, Spec (Design + Structure), Plan, Implement — one self-contained artifact per phase under thoughts/<task-id>/, each phase a fresh session reading only that artifact. Use it to run a ticket through qrspi, to research-then-plan-then-implement work too big for one session, to set up or advance thoughts/ artifacts, tell which phase one is in, or review one before moving on, and when a session drifts past 40% context and needs comp...Votes: 0GitHub stars: 2
- Token EfficiencyDiagnose and cut token consumption in agentic coding workflows — context rot, the 40% rule, intentional compaction, subagents as context firewalls, effort and model selection, prompt-caching invalidation, tool-output hygiene, KPIs. Use when asked why an agent is expensive or slow, when a session is drifting or re-exploring files it already read, when caching is not hitting (cache_read_input_tokens 0 on an unchanged prefix), when MCP tool schemas eat the window in /context, when choosing effor...Votes: 0GitHub stars: 2