Enforce Complete Code and Deliverable Generation

Enforces complete, production-critical output whenever a user asks for full files, complete implementations, exhaustive lists, or unabridged deliverables - treating a partial output as a broken one and optimizing for completeness over brevity. It bans specific patterns as hard failures: code-block placeholders (// ..., // TODO, // rest of code, // similar to above, bare ...), prose hedges ("let me know if you want me to continue," "for brevity," "and so on"), and structural shortcuts (skeleton output for a full-implementation request, showing first and last sections while skipping the middle, describing code instead of writing it).

[PAUSED - X of Y complete. Send "continue" to resume from: next section name]

Its execution process: scope the request by counting distinct deliverables and locking that number, build every deliverable completely with no partial drafts, then cross-check the final output's deliverable count against the locked scope count before responding. When a response approaches the token limit, it writes at full quality up to a clean breakpoint - end of a function, file, or section - rather than compressing or skipping ahead, and ends with a structured pause marker so "continue" resumes exactly where it stopped with no recap or repetition.

Use it when a user explicitly asks for full files, complete implementations, exhaustive lists, or unabridged deliverables, or when placeholder code or skipped sections would break the request. It does not override token limits, safety constraints, missing source context, or user-provided scope boundaries - and it explicitly forbids inventing unavailable code, credentials, private APIs, or project files just to appear complete.

Input: a request for a full file, complete implementation, or exhaustive deliverable set. Output: every requested deliverable in full, or, if the token limit is reached, a clean-breakpoint partial output ending in a structured [PAUSED - X of Y complete] marker for resumption.

Users and agents who need guaranteed-complete output - no placeholders, no "left as an exercise," no described-but-unwritten code - especially across long or multi-file requests.

Its final quick-check gate asks four things before any response is finalized: no banned pattern appears anywhere in the output, every requested item is present and finished, code blocks contain actual runnable code rather than descriptions of what code would do, and nothing anywhere in the response was shortened to save space. When a long output must split into labeled continuation chunks, each chunk is still expected to connect cleanly to the one before it, with no gap or repeated content at the seam.