forge substrate composes the rest into one
verdict; each stage is also callable on its own. See
The pre-action gate for the pipeline.
forge substrate
One pre-action gate: assumptions, route, impact, scope, memory, verify.
okToProceed:false, ask the returned assumption.questions before editing.
forge preflight
Assumption check — what a task names that the repo doesn’t define.
forge route
Recommend the cheapest capable model for a task.
forge impact
Hazard-aware blast radius — SCC-aware propagation and data-driven threshold from
PageRank centrality and ledger incident history. --basic reverts to fixed-threshold
mode.
forge collide
Parallel-session conflict radar: which recent teammate/agent sessions touched the
files (or their import neighbors) you are editing, read from the team-merged ledger —
risk = 1 − ∏(1 − rec × strength) over recent foreign sessions. Also exposed as the
collide_check MCP tool.
forge rank
Load-bearing code: weighted PageRank centrality over the atlas graph joined with
past-incident history from the evidence ledger (hazard = centrality × 1+history),
plus circular-import clusters (Tarjan SCC) and chokepoint files (articulation points).
Also exposed as the rank_code MCP tool.
forge scope
Decompose files into independent clusters — plus coupled files you didn’t name.
forge context
Budgeted context assembly + completeness gate — what an edit NEEDS known.
forge anchor
Goal-drift check — are your actual (git) changes still on the stated goal?
forge diagnose
Doom-loop check — record a failure; the same signature 3× mints a diagnosis + escalation.
forge imagine
Consequence simulation — predicted breaks + the minimal dry-run test suite for a task.
forge lean
Scope-minimality (M5) — measure the diff’s footprint vs what the task asked for.
The
route, impact, scope, context, anchor, and lean stages all run inside
forge substrate. Call them individually when you want just one signal.