agentii.tasks

Task identity (Q79)

One task = ticker × skill × 单个 mode. --mode=all is a generation-time expansion — it never exists as a single task (matching validate-mode-syntax.py's all reserved rule). Pre-M1: mode is default for the 53/62 skills without references/modes.md; Depth-Tier Light falls back to the tier's explicit skill list until essentials_modes exists.

Format (plan contract)

- [ ] T### [P] [pillar] TICKER × SKILL × MODE — purpose (src: pillar-id)

• [P] = different files and no incomplete deps; _cross/ synthesis tasks are never [P].
• Every task carries a src: ref — traceability survives into converge (Q26).
• tasks.md is strictly append-only thereafter.

Invocation

# Resolve the kit root first — contracts/kit-root.md. The kit's scripts do NOT ship
# with the skills, so never assume the CWD is the checkout.
KIT=""
for c in "${AGENTII_KIT_ROOT:-}" \
         "$(cat "$HOME/.claude/skills/agentii/.kit-root" 2>/dev/null)" \
         "${CLAUDE_PLUGIN_ROOT:-}"; do
  [ -n "$c" ] && [ -f "$c/scripts/agentii_cmd.py" ] && { KIT="$c"; break; }
done
if [ -z "$KIT" ]; then d="$PWD"; while [ "$d" != "/" ]; do
  [ -f "$d/scripts/agentii_cmd.py" ] && { KIT="$d"; break; }; d="$(dirname "$d")"; done; fi
[ -n "$KIT" ] || { echo "agentii kit not found — see contracts/kit-root.md" >&2; exit 1; }

python3 "$KIT/scripts/agentii_cmd.py" tasks --thesis theses/001-ai-semiconductors --spec theses/001-ai-semiconductors/spec.md

Gate 3 (after tasks, informational) shows task count × cost estimate.