What causes "cooking formula: %w" in gastownhall/beads?

Inspect the wrapped inner error to identify whether it is a storage or formula-content failure

cooking formula: %w

Error message

cooking formula: %w

What it means

After (optionally) clearing an old proto, persist-mode cook calls `cookFormula` to materialize the formula into a new proto bead. Any failure there — store write errors, step instantiation failures, dependency creation problems — is wrapped as `cooking formula`. The proto creation itself, not loading or resolving, failed at this point.

Source

Thrown at cmd/bd/cook.go:328

// persistCookFormula creates a proto bead in the database (persist mode)
func persistCookFormula(ctx context.Context, resolved *formula.Formula, protoID string, force bool, vars, bondPoints []string) error {
	// Check if proto already exists
	existingProto, err := store.GetIssue(ctx, protoID)
	if err == nil && existingProto != nil {
		if !force {
			return fmt.Errorf("proto %s already exists (use --force to replace)", protoID)
		}
		// Delete existing proto and its children
		if err := deleteProtoSubgraph(ctx, store, protoID); err != nil {
			return fmt.Errorf("deleting existing proto: %w", err)
		}
	}

	// Create the proto bead from the formula
	result, err := cookFormula(ctx, store, resolved, protoID)
	if err != nil {
		return fmt.Errorf("cooking formula: %w", err)
	}

	if jsonOutput {
		return outputJSON(cookResult{
			ProtoID:    result.ProtoID,
			Formula:    resolved.Formula,
			Created:    result.Created,
			Variables:  vars,
			BondPoints: bondPoints,
		})
	}

	fmt.Printf("%s Cooked proto: %s\n", ui.RenderPass("✓"), result.ProtoID)
	fmt.Printf("  Created %d issues\n", result.Created)
	if len(vars) > 0 {
		fmt.Printf("  Variables: %s\n", strings.Join(vars, ", "))
	}
	if len(bondPoints) > 0 {

View on GitHub (pinned to 71377f2769)

Solutions

Inspect the wrapped inner error to identify whether it is a storage or formula-content failure
Run `bd doctor` to verify database health and repair common issues
Validate the formula (`bd formula validate`) to rule out malformed steps before re-running
Retry after resolving the storage issue; if the run partially wrote beads, clean up the partial proto before re-running with `--force`

Defensive patterns

Strategy: try-catch

Validate before calling

// validate formula before persisting
if err := formula.Validate(resolved); err != nil {
    return fmt.Errorf("formula invalid, refusing persist: %w", err)
}

Try / catch

result, err := cookFormula(ctx, store, resolved, protoID)
if err != nil {
    // partial proto may exist; clean up before retry
    _ = deleteProtoSubgraph(ctx, store, protoID)
    return fmt.Errorf("cooking formula: %w", err)
}

Prevention

Run bd doctor / verify store writability before persist runs in CI
Validate formulas before cooking; avoid cooking with a degraded or locked database
Serialize persist-mode cooks so concurrent writers can't conflict

When it happens

Trigger: `bd cook --mode=persist` where `cookFormula` fails: the store rejects bead creation (duplicate keys, DB unavailable), a step fails to instantiate, or dependency edges between created steps cannot be written.

Common situations: Dolt/database connection problems during write; formula producing invalid steps or references; resource constraints (disk full); concurrent writers conflicting in the same beads database.

Related errors

AI-assisted analysis of gastownhall/beads@71377f2769 (2026-08-30). Data as JSON: /api/errors/d8fe8181252f37e4. Report an issue: GitHub.