Como usar
- Copie o comando e verifique o pacote antes de instalar no seu ambiente de agente.
- Consulte no original as ferramentas, credenciais e ambientes necessários.
- Descreva os dados de entrada e a entrega esperada; confira a saída com as fontes.
npx skills add https://github.com/gasserane/personal-skills --skill implementation-packAntes de começar
- A instalação não inclui assinatura de modelo, franquia API ou acesso aos seus arquivos.
- Estrelas e instalações indicam atividade, não um teste independente de qualidade ou segurança.
Renderizado do repositório de origem, preservando títulos, exemplos, código, tabelas, links e imagens.
/implementation-pack — post-award PM pack generator
You generate a standalone, IPPF-branded implementation pack a full project manager runs with no AI dependency. You are the front-half: load baseline, apply awarded figures, build. Vi executes any optional reporting spawns.
Arguments
--donor <name>(defaultcerv). Maps toane_package/proposals/donor_profiles/<name>.json.- Optional baseline path (a proposal
handback.jsonor a proposal pack folder).
Step 1 — Load the donor profile
from ane_package.proposals.config import load_profile
profile = load_profile("cerv") # or the --donor valueState the implementation config back to Ane: reporting periods, RAG thresholds.
Step 2 — Load the baseline (never invent)
from ane_package.proposals.baseline_io import load_baseline
baseline = load_baseline(handback_path_or_dict)If no baseline is available, ask Ane for the proposal handback or a baseline brief. Do not invent any WP, indicator, figure, or partner (factual-reliability rule).
Step 3 — Apply the awarded figures
Ask Ane for the signed grant-agreement budget, dates, and revised targets.
from ane_package.proposals.baseline_io import apply_awarded
baseline = apply_awarded(baseline, awarded) # raises BaselineError on unknown WP/indicator or incoherenceIf the project is not yet awarded, skip this step; the pack flags "baseline = submitted" in the Guide and README.
Step 4 — Build the pack
from ane_package.proposals.implementation_pack import build_implementation_pack
manifest = build_implementation_pack(profile, baseline, out_dir)Apply mel_wiki/wiki/concepts/edit-preservation-protocol.md when target file exists — if the output folder already holds an edited pack, treat Ane's content as canonical and edit scope-bounded, do not regenerate from scratch.
Step 5 — Optional reporting cycle (only when Ane asks for a populated report)
Delegate to Vi to spawn mel-report-writer (draft the narrative from tracker status) and contribution-plausibility-analyst (judge whether indicator progress supports the outcome claims). Neither is needed for the pack to be usable.
Step 6 — Disclosure and scope boundary
- The Word report carries the AI-disclosure colophon per
mel_wiki/wiki/concepts/ai-use-in-publications.md. AI is never an author. - State the scope boundary: the finance system of record, the official portal/grant-management reporting, and the authorising officer's sign-off are owned by finance, the portal, and the authorising officer — not by this skill.
Output
Return the pack folder path and the manifest. Surface whether the baseline is awarded or submitted, and any [PM: insert X] the PM must still complete.
