Functionality pruner

This skill governs decisions about whether functionality justifies its existence. It runs in two stages: a necessity gate (does the problem this code addresses actually occur in this context?) followed by a worth ledger (does the value justify the cost?). It applies equally to unimplemented features (accept / reject / minimize) and to existing code (keep / simplify / delete / remove-as-obsolete). For measuring complexity itself, see structural-simplification. For the upstream principles (YAGNI, scope control, proportional solutions), see architecture-guidelines.

Core Directives

1. Necessity precedes worth. Before scoring value and cost, verify the problem the code addresses can actually occur in this context. Code guarding against architecturally impossible states has no product value for that failure mode. Skip the worth ledger and emit OBSOLETE unless the code is serving as the canonical executable invariant (§1c).
2. Separate the ledger. Value and cost are distinct axes. Score each independently; never collapse into a single number.
3. Cost compounds, value decays. Value is realized per use; cost accrues on every future change, test run, review, and incident. Always evaluate over the feature's expected lifetime.
4. The default is No. If worth is not clearly positive, reject or minimize. YAGNI is the null hypothesis.
5. Build and audit share a model. The same axes apply whether deciding what to add or what to remove. A feature that would fail as a proposal today should fail as existing code today.
6. Remove over refactor, refactor over rewrite. A retrospective audit that finds negative worth — or that fails the necessity gate — prefers safe removal or deprecation to elaborate justification. Removal still follows migration, rollback, and compatibility constraints.
7. Outcome evidence informs worth; it is not the verdict. Consume current linked outcome evidence when available. Completion, deployment, or adoption alone cannot prove downstream value, and no hypothesis state automatically dictates a worth decision.

---

1. The Necessity Gate

The Worth Model (§2) assumes the code under review is solving a real problem. Before scoring V and C, confirm that the problem itself exists in this stack. If it does not, the worth ledger does not apply: emit OBSOLETE in retrospective mode, or DROP with a necessity-failure rationale in prospective mode. For retrospective removals, apply the safety constraints in §7b before changing code.

[!IMPORTANT] A monorepo single-page application deployed as one artifact cannot run client and server at different versions; a "client version check" in that stack guards against an impossible state. It has no V for that failure mode — not low V — because the failure mode it prevents cannot occur. Worth scoring would mis-classify this as low-V / low-C "DEFER" or "KEEP." The necessity gate catches it.

1a. Categories of non-problem-solving code

Category	Definition	Typical example
Impossible-state guard	Defends against a state ruled out by deployment topology, type system, or runtime invariant	Client/server version skew in a single-artifact SWA; null-guard on a non-nullable type; race-condition mutex in a single-threaded executor; retry loop on a deterministic in-process call
Already-defended-elsewhere	Concern fully owned by a different layer, duplicated here	XSS-escaping atop a templating engine that already escapes; manual rollback inside an outer transaction; CSRF token on an idempotent GET; HTTPS-upgrade logic when the load balancer terminates TLS
Cargo-culted pattern	Pattern whose prerequisites do not hold in this context	Connection pool in a CLI that exits in 200 ms; singleton in a stateless lambda; client-side request dedupe against an idempotent endpoint; back-compat shim for a client class that no longer exists
Phantom requirement	Solves a requirement that was never real or has lapsed	Feature flag for a completed launch; A/B branch after the experiment concluded; migration code that has provably run on every record
Generality without instantiation	Abstraction whose anticipated variation never materialized	Strategy pattern with one strategy; plugin interface with one implementation; config key that has held one value across all environments for the feature's lifetime
Logically dead branch	Branch unreachable given upstream contracts	if (!user.id) after auth middleware that guarantees it; try/catch around statically non-throwing code; default values for parameters callers always populate

1b. Detection heuristics

Run these BEFORE scoring V or C. A high-confidence positive result routes the verdict to OBSOLETE (retrospective) or DROP-as-non-problem (prospective), subject to the invariant-documentation and load-bearing exceptions in §§1c/8e.

Heuristic	Signal	Catches
Invariant audit	List the invariants the architecture, type system, deployment topology, and trust boundary maintain. List the conditions the code branches on. Branches that contradict an invariant are dead.	Impossible-state guards, dead branches
Trigger reachability	Construct a concrete real-world sequence that activates the code without violating an architectural invariant. Failure to construct one after checking callers, entry points, tests, and runtime paths is a positive finding.	Impossible-state guards, dead branches
Origin archaeology	Pull the introducing commit / PR / ADR. Verify the rationale's premises still hold (dependency present, platform supported, client class extant, migration incomplete). Lapsed premises mean the code is obsolete.	Phantom requirements
Layer-responsibility map	For each cross-cutting concern (auth, escaping, retry, validation, caching), name the single layer that owns it. Other layers performing the same job are redundant or signal a missing trust boundary.	Already-defended-elsewhere
Pattern-prerequisite check	For each recognizable pattern, list its prerequisites (long-lived process, mutable shared state, non-idempotent dependency, multiple implementations). Prerequisites that do not hold here mean the pattern is cargo-culted.	Cargo-culted patterns
One-value config	A flag, env var, or config key that has held one value across all environments for the feature's lifetime is a dead-seam candidate. Either inline the value or document the concrete second value, compliance requirement, or pending rollout that keeps it alive.	Generality without instantiation, phantom reqs.
Zero-everything signature	Production code with zero telemetry hits AND zero bug history AND zero recent edits is not necessarily "stable" — it may have never run. Combine with the invariant audit to distinguish load-bearing-but-quiet from guarding-the-impossible.	Impossible-state guards

[!IMPORTANT] The invariant audit is the highest-yield necessity check. Most non-problem-solving code is defending against violations of invariants the surrounding stack already guarantees. Enumerate those invariants explicitly before reading the code, then walk the branches with the list in hand.

1c. Necessity findings that are not deletions

[!WARNING] Some "impossible-state" code is documenting an invariant rather than enforcing one — an assert version_match whose purpose is to fail loudly if a future contributor changes the deployment topology. That has small but real value as machine-checkable documentation. The fix is usually to convert it to a comment, an ADR reference, a build-time check, or a test — not silent deletion. If the code is the canonical record of an invariant nothing else captures, SIMPLIFY (downgrade to documentation) rather than OBSOLETE.

Inverse failure mode: see §8e. Some complexity that resembles cargo-culting or over-engineering is in fact load-bearing because the simple version was measured to be too slow, too unsafe, or too fragile. Read the original rationale before voting OBSOLETE on anything that merely looks like a non-problem.

1d. Necessity vs. low worth

The distinction matters for the audit record.

Verdict	Rationale	Future re-litigation risk
OBSOLETE	"The problem this code addresses cannot occur in this stack."	Low — the rationale is structural; only an architecture change reopens it.
DELETE	"The value does not justify the cost."	Higher — priorities or cost shift and the case reopens.

Record the distinction so a later audit does not reintroduce the same code under new conditions. "We removed the client version check because it cost more than it returned" invites debate about thresholds; "we removed it because client/server version skew cannot occur in a single-artifact deploy" closes the question.

1e. Obligation vs. mechanism

A subject can pass the necessity gate — the problem is real — while its grain is still inflated by mechanism nobody demanded. Before scoring worth, restate each requirement behind the subject in two parts:

• Obligation — the outcome, evidence, or restriction that must exist (a record with actor/time, a gate before a step, an actor limitation).
• Mechanism — the specific rights, roles, endpoints, record types, or protocols the requirement text or the implementation chose to satisfy it.

Mechanism the obligation does not force is a SIMPLIFY candidate even when the functionality itself is KEEP. Audit in two passes: first within the current requirements to establish the floor, then treat careful requirement edits as prospective candidates scored on this same ledger. Flag edits with real external trade-offs (consent models, public intake, protocol surfaces) as explicit product decisions rather than deciding them silently, and route changes of requirement meaning through requirements-grounding. Three floors are never negotiable: legal/regulatory obligations, separation-of-duties (second-person) controls, and external protocol surfaces others depend on.

Typical yields: an actor condition encoded as a dedicated right or role where a membership attribute or workflow-state gate satisfies the same acceptance criterion; a person-split where the obligation only demands recorded evidence before the next step; one endpoint per read projection of an aggregate the caller already fetches.

---

2. The Worth Model

Once the necessity gate (§1) passes, the question becomes: does the value delivered justify the cost imposed? Worth is the relation between Value (V) delivered and Cost (C) imposed over lifetime L. Both sides are multi-dimensional.

Value axes

Axis	Symbol	What it measures	Measurability
Utility	U	Severity of the user need; what actually breaks without it	Judgment, user research
Frequency	F	How often the need arises per affected user per unit time	Measurable (telemetry)
Reach	R	Proportion of users / flows / environments that encounter the need	Measurable (analytics)
Irreplaceability	I	Cost of the next-best alternative (workaround, external tool, doing without)	Judgment, comparative

Aggregate product value ≈ U × F × R × I. If any axis is zero, ordinary product value is zero; external floors, keystone cost, and safety exceptions are handled separately in §8.

[!IMPORTANT] A feature loved by 2% of users, used once a year, with a trivial workaround, has near-zero total value no matter how elegant it is. Score honestly — especially R and F, which are routinely inflated.

Cost axes

Structural cost is delegated to structural-simplification: the Component-kinds Δ, Dependency-edges Δ, Max-chain-depth Δ, Module-count Δ introduced (prospective) or already present (retrospective). See the Reporting Vocabulary in structural-simplification for the symbol mapping. This skill adds three ongoing-cost axes that structure alone does not capture:

Axis	Symbol	What it measures	Measurability
Maintenance	M	Tests, docs, reviews, dependency updates the feature demands	Measurable (test/doc count, churn)
Risk	X	Bug surface × blast radius; security, privacy, performance exposure	Measurable (defect history, incidents)
Evolution tax	E	Degree to which the feature constrains future change	Judgment, changelog trace

Aggregate cost over lifetime, in axis-symbol form (one-time structural delta plus ongoing maintenance × lifetime):

Aggregate cost ≈ (ΔD + ΔK + ΔP + Δn) + (M + X + E) × L

— where ΔD, ΔK, ΔP, Δn are the structural deltas from structural-simplification (see its Reporting Vocabulary: Component-kinds Δ, Dependency-edges Δ, Max-chain-depth Δ, Module-count Δ).

The worth inequality

Worth > 0   ⇔   V × L   >   C_structural + (M + X + E) × L

For short-lived code, the structural footprint dominates. For long-lived code, M + X + E dominates. Most production features are long-lived; plan for the ongoing term.

[!WARNING] Evolution tax (E) is the most-underestimated axis because it is invisible in the current code review. It shows up later, as the PR that "should have been small but touched twelve files."

---

3. Two Modes

The model is the same; the inputs differ. Both modes run the necessity gate (§1) before scoring worth.

3a. Prospective — evaluating proposed functionality

Applied to tickets, specs, PRDs, loose ideas, or PR scope before implementation. All inputs are estimates; record confidence explicitly.

1. State the functionality in one sentence: "This allows [who] to [do what] so that [outcome]."
2. Run the necessity gate (§1). Confirm the failure mode addressed is reachable in the target stack and is not already owned by another layer. A prospective necessity failure is rare but consequential: it stops a build that would have produced dead code on day one.
3. Score V axes with evidence: user interviews, request tickets, analytics of the workaround, competitor behavior. Opinions are not evidence. Unsupported opinions are not enough evidence for high-confidence build decisions. Cite linked outcome hypotheses when present; before release they express expected value, not observed impact.
4. Score C axes against a concrete implementation sketch: files touched, new abstractions or dependencies introduced, tests required, failure modes created.
5. Apply the Decision Protocol (§6). Verdicts are prospective (§7a).

3b. Retrospective — auditing existing functionality

Applied to code, modules, features, capabilities, or flags that already exist. Inputs are observable; bias toward measurement over judgment.

1. Define the boundary: files, symbols, entry points, feature flags, routes, or callers.
2. Run the necessity gate (§1). Walk the heuristics in §1b before any worth scoring. A positive finding short-circuits the rest of the audit to OBSOLETE.
3. Score V from usage data:
   • Telemetry hits per time window, per user cohort.
   • Reach: unique users or flows that enter this code path.
   • Irreplaceability: does an alternative path exist? Do users already use it?
   • If V cannot be measured, that itself is a finding — instrument, identify an external floor (§8c), or keep confidence Low.
   • Consume current outcome-evidence records from requirements-traceability. Preserve the canonical hypothesis version, cohort, threshold, window, and guardrails. stale or inconclusive evidence cannot support High value confidence; rejected evidence lowers the supported value claim but does not by itself prove zero value.
4. Score C from current observable state:
   • Structural: measure D, K, P, n per structural-simplification.
   • M: dedicated tests, doc pages, recent commit churn, dependency drift.
   • X: bug ticket history, incident postmortems, security/perf hotspot reports.
   • E: count of PRs / design docs where this feature caused scope expansion, workarounds, or delays.
5. Apply the Decision Protocol (§6). Verdicts are retrospective (§7b).

[!NOTE] A retrospective audit with no telemetry available should first return an instrumentation task, not a verdict — unless the necessity gate has already produced a finding, in which case telemetry is not needed (you cannot measure usage of a code path that cannot be triggered). Deciding to delete a feature merely because you cannot see it being used is survivorship bias in reverse; deciding to remove it because the failure mode it guards against cannot occur is structural reasoning.

---

4. Heuristic Checks

Fast worth signals — usage silence, workaround in the wild, single caller, flag defaulted off, orphan test, churn hotspot, churn × complexity, defect clustering, bug-fix-to-feature ratio, blocked PRs, documentation rot — and the axis each one moves. The necessity heuristics in §1b run first. Read references/worth-signals.md when scoring V or C in retrospective mode, and run churn × complexity before any subjective judgment.

---

5. Forcing Questions

Four interrogations — necessity, value, cost, counterfactual — each exposing a common failure mode. Answers MUST be written, not implicit. Read references/forcing-questions.md and answer the necessity questions before any value scoring. A removal cost in 12 months that exceeds the build cost today is a one-way door: apply §8 before committing.

---

6. Decision Protocol

1. Run the necessity gate (§1). Walk the heuristics in §1b. If the code addresses a problem that cannot occur in this context, emit OBSOLETE (retrospective) or DROP with a necessity-failure rationale (prospective). Skip remaining steps.
2. Score V axes (U, F, R, I) on a 0–3 scale with one-line evidence per axis. When outcome evidence exists, cite its hypothesis ID, state, freshness, and observation identity; do not replace its threshold or guardrails with a more favorable interpretation.
3. Score C axes:
   • Delegate D, K, P, n to structural-simplification (deltas for prospective; absolute measured values for retrospective).
   • Score M, X, E on 0–3 with one-line evidence per axis.
4. Record confidence (Low / Medium / High) for each side independently.
5. Compare across both ledgers without summing.
6. Classify using the Worth Matrix (§6a) and apply the confidence gate (§6b).
7. Emit the Output Contract (§9).

6a. The Worth Matrix

	Low C	Medium C	High C
High V	BUILD / KEEP	BUILD / KEEP	NEGOTIATE (§8)
Medium V	BUILD-minimal / KEEP	BUILD-minimal / SIMPLIFY	DEFER / SIMPLIFY
Low V	DEFER / QUARANTINE	DROP / SIMPLIFY	DROP / DELETE

Read the matrix identically in both modes. Prospective verdicts are accept / reject; retrospective verdicts are keep / simplify / delete. The matrix only applies when the necessity gate (§1) has passed; necessity failures bypass it entirely.

6b. Confidence gate

A verdict carries the confidence of its weakest input. If either V or C confidence is Low:

• Prospective → default to DEFER. Gather evidence before committing to high-cost action.
• Retrospective → default to QUARANTINE. Add instrumentation, revisit after N weeks with measured data.

Do not commit to irreversible verdicts (BUILD, DELETE) on low-confidence estimates. OBSOLETE is exempt from the confidence gate when the necessity finding is itself High confidence — a structural impossibility does not become more or less impossible with more data.

An unmeasured, inconclusive, or stale outcome assessment keeps the affected value claim Low unless independent current evidence supports it. supported may raise confidence only within the measured cohort, window, and guardrails. Authoritative floors in §8c remain source-driven and do not require an empirical outcome hypothesis.

---

7. Verdicts

7a. Prospective verdicts