nen-critical-path-forge

Critical Path Forge · 硬

Concentrate effort only after identifying the decisive point. Trade breadth for depth consciously, then restore broad verification before claiming success.

Qualify the target

Use this ability only when:

• evidence identifies one blocker or hypothesis as outcome-dominant;
• the intended action is inside current authority;
• the change is reversible or has an explicit recovery path;
• a hard budget and stop condition can be named; and
• adjacent surfaces can retain a minimum verification reserve.

If the target is still uncertain, use a perimeter or concealed-trace workflow first.

Perform the wind-up

1. State the exact target and expected result.
2. Record the evidence that makes it critical.
3. Define the one focused action or artifact.
4. Name the exposed surfaces receiving no implied assurance.
5. Reserve one rollback route and one post-strike integration check.
6. Set a cap for time, attempts, files, or tool calls using observable units.

The visible wind-up is an interface: it lets collaborators challenge the wrong target before concentrated work begins.

Strike once, learn once

• Make the smallest decisive change or experiment.
• Avoid opportunistic cleanup outside the target.
• Run the most discriminating focused proof first.
• On failure, update the hypothesis before another attempt.
• Never auto-repeat an ambiguous mutation.
• Stop when the cap, authority, or target assumption breaks.

Return to Ken

After the focused proof:

1. Run the minimum broad checks needed to catch adjacent regression.
2. Review the diff or artifact for unintended scope.
3. Restore any temporary instrumentation or narrowed configuration.
4. Report the target, strike, proof, and exposed surface separately.

Vow

Maximum focus means reduced coverage elsewhere. Never translate a focused pass into system-wide confidence. Never use speed to justify an irreversible, external, secret-bearing, destructive, purchasing, publishing, or messaging effect that the task did not authorize.