teleo-codex/domains/ai-alignment/the determinism boundary separates guaranteed agent behavior from probabilistic compliance because hooks enforce structurally while instructions degrade under context load.md

type	domain	secondary_domains	description	confidence	source	created	depends_on	challenged_by	related	reweave_edges
claim	ai-alignment	collective-intelligence	Agent behavior splits into two categories — deterministic enforcement via hooks (100% compliance) and probabilistic guidance via instructions (~70% compliance) — and the gap is a category difference not a performance difference	likely	Cornelius (@molt_cornelius), 'Agentic Systems: The Determinism Boundary' + 'AI Field Report 1' + 'AI Field Report 3', X Articles, March 2026; corroborated by BharukaShraddha (70% vs 100% measurement), HumanLayer (150-instruction ceiling), ETH Zurich AGENTbench, NIST agent safety framework	2026-03-30	iterative agent self-improvement produces compounding capability gains when evaluation is structurally separated from generation	AI integration follows an inverted-U where economic incentives systematically push organizations past the optimal human-AI ratio	trust asymmetry between agent and enforcement system is an irreducible structural feature not a solvable problem because the mechanism that creates the asymmetry is the same mechanism that makes enforcement necessary	trust asymmetry between agent and enforcement system is an irreducible structural feature not a solvable problem because the mechanism that creates the asymmetry is the same mechanism that makes enforcement necessary|related|2026-04-03

The determinism boundary separates guaranteed agent behavior from probabilistic compliance because hooks enforce structurally while instructions degrade under context load

Agent systems exhibit a categorical split in behavior enforcement. Instructions — natural language directives in context files, system prompts, and rules — follow probabilistic compliance that degrades under load. Hooks — lifecycle scripts that fire on system events — enforce deterministically regardless of context state.

The quantitative evidence converges from multiple sources:

• BharukaShraddha's measurement: Rules in CLAUDE.md are followed ~70% of the time; hooks are enforced 100% of the time. The gap is not a performance difference — it is a category difference between probabilistic and deterministic enforcement.
• HumanLayer's analysis: Frontier thinking models follow approximately 150-200 instructions before compliance decays linearly. Smaller models decay exponentially. Claude Code's built-in system prompt already consumes ~50 instructions before user configuration loads.
• ETH Zurich AGENTbench: Repository-level context files reduce task success rates compared to no context file, while increasing inference costs by 20%. Instructions are not merely unreliable — they can be actively counterproductive.
• Augment Code: A 556:1 copy-to-contribution ratio in typical agent sessions — for every 556 tokens loaded into context, one meaningfully influences output.
• NIST: Published design requirement for "at least one deterministic enforcement layer whose policy evaluation does not rely on LLM reasoning."

The mechanism is structural: instructions require executive attention from the model, and executive attention degrades under context pressure. Hooks fire on lifecycle events (file write, tool use, session start) regardless of the model's attentional state. This parallels the biological distinction between habits (basal ganglia, automatic) and deliberate behavior (prefrontal cortex, capacity-limited).

The convergence is independently validated: Claude Code, VS Code, Cursor, Gemini CLI, LangChain, and Strands Agents all adopted hooks within a single year. The pattern was not coordinated — every platform building production agents independently discovered the same need.

Additional Evidence (supporting)

The habit gap mechanism (AN05, Cornelius): The determinism boundary exists because agents cannot form habits. Humans automatize routine behaviors through the basal ganglia — repeated patterns become effortless through neural plasticity (William James, 1890). Agents lack this capacity entirely: every session starts with zero automatic tendencies. The agent that validated schemas perfectly last session has no residual inclination to validate them this session. Hooks compensate architecturally: human habits fire on context cues (entering a room), hooks fire on lifecycle events (writing a file). Both free cognitive resources for higher-order work. The critical difference is that human habits take weeks to form through neural encoding, while hook-based habits are reprogrammable via file edits — the learning loop runs at file-write speed rather than neural rewiring speed. Human prospective memory research shows 30-50% failure rates even for motivated adults; agents face 100% failure rate across sessions because no intentions persist. Hooks solve both the habit gap (missing automatic routines) and the prospective memory gap (missing "remember to do X at time Y" capability).

Additional Evidence (supporting)

7 domain-specific hook implementations (Cornelius, How-To articles, 2026): Each domain independently converges on hooks at the point where cognitive load is highest and compliance most critical:

1. Students — session-orient hook: Loads prerequisite health and upcoming exam context at session start. Fires before the agent processes any student request, ensuring responses account for current knowledge state.
2. Fiction writers — canon gate hook: Fires on every scene file write. Checks new content against established world rules, character constraints, and timeline consistency. The hook replaces the copy editor's running Word document with a deterministic validation layer.
3. Companies — session-orient + assumption-check hooks: Session-orient loads strategic context and recent decisions. Assumption-check fires on strategy document edits to verify alignment with stated assumptions and flag drift from approved strategy.
4. Traders — pre-trade check hook: Fires at the moment of trade execution — when the trader's inhibitory control is most degraded by excitement or urgency. Validates the proposed trade against stated thesis, position limits, and conviction scores. The hook externalizes the prefrontal discipline that fails under emotional pressure.
5. X creators — voice-check hook: Fires on draft thread creation. Compares the draft's voice patterns against the creator's established identity markers. Prevents optimization drift where the creator unconsciously shifts voice toward what the algorithm rewards.
6. Startup founders — session-orient + pivot-signal hooks: Session-orient loads burn rate context, active assumptions, and recent metrics. Pivot-signal fires on strategy edits to check whether the proposed change is a genuine strategic pivot or a panic response to a single data point.
7. Researchers — session-orient + retraction-check hooks: Session-orient loads current project context and active claims. Retraction-check fires on citation to verify the cited paper's current status against retraction databases.

The pattern is universal: each hook fires at the moment where the domain practitioner's judgment is most needed and most likely to fail — execution under emotional load (traders), creative flow overriding consistency (fiction), optimization overriding authenticity (creators), urgency overriding strategic discipline (founders). The convergence across 7 unrelated domains corroborates the structural argument that the determinism boundary is a category distinction, not a performance gradient.

Challenges

The boundary itself is not binary but a spectrum. Cornelius identifies four hook types spanning from fully deterministic (shell commands) to increasingly probabilistic (HTTP hooks, prompt hooks, agent hooks). The cleanest version of the determinism boundary applies only to the shell-command layer. Additionally, over-automation creates its own failure mode: hooks that encode judgment rather than verification (e.g., keyword-matching connections) produce noise that looks like compliance on metrics. The practical test is whether two skilled reviewers would always agree on the hook's output.

---

Relevant Notes:

• iterative agent self-improvement produces compounding capability gains when evaluation is structurally separated from generation — the determinism boundary is the mechanism by which evaluation separation is enforced: hooks guarantee the separation, instructions merely suggest it
• coding agents cannot take accountability for mistakes which means humans must retain decision authority over security and critical systems regardless of agent capability — the determinism boundary provides a structural mechanism for retaining decision authority through hooks on destructive operations

Topics:

• _map