pm-archive

Preamble (run first)

source ~/.nanopm/lib/nanopm.sh 2>/dev/null || \
  source .nanopm/lib/nanopm.sh 2>/dev/null || \
  { echo "ERROR: nanopm not installed. Run: curl -fsSL https://raw.githubusercontent.com/nmrtn/nanopm/main/setup | bash"; exit 1; }
nanopm_preamble

What this skill does

Archives one entity — writes status: archived and a polished archived_reason sentence to its frontmatter, then re-indexes (if an opportunity) and pushes the change to Supabase.

The raw reason from the user is always rephrased by the agent into one specific, contextual sentence that references the actual feature, PR, or decision. The user's wording is never stored verbatim.

Kind → opening phrase mapping:

• implemented → "Implemented via …"
• dropped → "Dropped — …"
• absorbed → "Absorbed into … — …"
• completed → "Completed — …"
• abandoned → "Abandoned — …"
• defunct → "Defunct — …"
• pivoted → "Pivoted — …"
• irrelevant → "No longer relevant — …"
• outgrown → "Outgrown — …"

Steps

1. Resolve the entity

Parse the argument (opportunity/cold-start-context → type=opportunity, slug=cold-start-context).

_ARG="${1:-}"
_TYPE="${_ARG%%/*}"
_SLUG="${_ARG#*/}"
[ -n "$_TYPE" ] && [ -n "$_SLUG" ] && [ "$_TYPE" != "$_SLUG" ] || {
  echo "ERROR: argument must be <type>/<slug>, e.g. opportunity/cold-start-context"
  exit 1
}
_ENTITY_DIR=".nanopm/wiki/entities/${_TYPE}s"
_ENTITY_FILE="$_ENTITY_DIR/$_SLUG.md"
[ -f "$_ENTITY_FILE" ] || {
  echo "ERROR: entity file not found: $_ENTITY_FILE"
  exit 1
}

2. Read the entity for context

Read $_ENTITY_FILE — you need the title, theme (if any), and first paragraph of the summary to write a specific, accurate reason.

3. Parse kind and raw detail from the launch context

The launch context (user context) contains the raw reason. Extract:

• kind — the first word before the colon (e.g. implemented, dropped, absorbed)
• detail — everything after the colon (e.g. shipped in Supabase sync PR)