v4.1.1

Merge pull request #22 from sh4i-yurei/fix/mcp-stdout-to-stderr

fix(mcp): redirect warnings to stderr to preserve MCP JSON integrity

v4.1.0

Added

• DRR Linking (FPF E.9): Decision records now create graph relations to hypotheses.

  • quint_decide accepts new rejected_ids parameter for rejected alternatives.
  • Creates selects relation: DRR → winner hypothesis.
  • Creates rejects relations: DRR → each rejected alternative.
  • Enables queries: "What alternatives were considered for this DRR?"

• Context Markdown Formatting: quint_record_context now normalizes input to proper markdown.

  • Vocabulary entries formatted as definition list (- **Term**: definition).
  • Invariants formatted as numbered list with line breaks.

• Audit Tree Shows MemberOf Relations: quint_audit_tree now displays memberOf groupings.

  • Decision context alternatives are shown under [members] section.
  • Member R scores displayed for comparison (no WLNK propagation per FPF spec).

• Improved Tool Descriptions: quint_propose parameters now have actionable descriptions.

  • depends_on: Explains WLNK impact and when to use.
  • decision_context: Explains grouping purpose and when to use.
  • q1-hypothesize.md: Added "Linking Checklist" before proposing hypotheses.

• Merged /q-audit into /q-query: Simplified command set.

  • /q-query now shows R_eff and dependency trees for found holons.
  • Removed standalone /q-audit utility (redundant with enhanced /q-query).
  • /q4-audit (FPF phase) remains unchanged.

• sqlc Integration: Type-safe database queries generated from SQL.

  • All database operations now use sqlc-generated code with proper type safety.
  • New db/store.go wrapper provides clean API while preserving schema bootstrap.
  • Added comprehensive test suite for database operations (db/store_test.go).

• GetHolon Query: Added query to fetch hypothesis metadata by ID (foundation for future Kind-CAL work).

• New MCP Tools for Trust Calculus (B.3):

  • quint_audit_tree: Visualize assurance tree with R scores, dependencies, and CL penalties.
  • quint_calculate_r: Compute R_eff with detailed breakdown (self score, weakest link, decay penalties).
  • quint_check_decay: Identify holons with expired evidence (epistemic debt detection).

• Parent ID Chain (FPF Enforcement):

  • Added parent_id foreign key to holons table for L0→L1→L2 progression tracking.
  • New queries: GetHolonsByParent, GetHolonLineage for traversing hypothesis chains.
  • CreateHolon now accepts parent_id parameter for linking hypothesis progression.
  • Enables auditable chain from L2 decision back to original L0 hypothesis.

• Derived Phase (FPF Enforcement):

  • Phase is now computed from holons.layer data instead of stored in state.json.
  • New DerivePhase() method computes phase from database state.
  • New GetPhase() returns derived phase when DB available, falls back to State.Phase.
  • Prevents AI bypass of FPF phase controls via direct file manipulation.

• Audit Logging (FPF Enforcement):

  • New audit_log table tracks all MCP tool invocations.
  • Captures: tool name, operation, actor, target ID, input hash, result, and details.
  • Instrumented tools: quint_propose, quint_verify, quint_decide, quint_move.
  • Enables detection of FPF bypasses through audit trail analysis.
  • Context-aware logging supports multi-session isolation.

• Self-Healing Signed Projections (FPF Enforcement):

  • All hypothesis/evidence/DRR files now include content_hash in YAML frontmatter.
  • New WriteWithHash() function adds cryptographic hash (SHA-256 truncated) on write.
  • New ValidateFile() detects tampering by comparing stored vs computed hash.
  • New ReadWithValidation() on Tools automatically detects and logs tampering.
  • When tampering detected: regenerates file from DB (DB is source of truth).
  • Tampering events logged to audit_log for violation tracking.

• DRR Holon Tracking:

  • FinalizeDecision now creates DRR holon in database (enables derived phase detection).
  • DRR holons linked to winner hypothesis via parent_id.

• Tool Preconditions (FPF Enforcement):

  • All MCP tools now validate preconditions before execution.
  • quint_propose: Validates title, content, and kind fields.
  • quint_verify: Confirms hypothesis exists in L0, validates verdict.
  • quint_test: Ensures hypothesis is in L1 (not L0), validates verdict.
  • quint_audit: Confirms hypothesis is in L2 before audit.
  • quint_decide: Requires L2 hypotheses exist, validates winner_id and title.
  • quint_calculate_r / quint_audit_tree: Validates holon existence in DB.
  • Precondition failures logged to audit_log with BLOCKED status.
  • Each error includes actionable suggestion for the user.

• Command Contracts (FPF Enforcement):

  • All FPF command prompts (q0-q5) now include formal enforcement contracts.
  • YAML frontmatter with pre, post, invariant, and required_tools fields.
  • RFC 2119 bindings (MUST, MUST NOT, SHALL) for mandatory behaviors.
  • "Invalid Behaviors" section explicitly lists protocol violations.
  • "Checkpoint" section with verification checklist before phase transition.
  • Success/failure path examples with few-shot learning.
  • "State machine executor" mechanical persona to reduce AI improvisation.
  • Defense in depth: soft gate (prompts) + hard gate (preconditions).

• Inline Schema Migrations:

  • Existing databases automatically upgraded on startup.
  • Adds parent_id and cached_r_score columns to existing holons table.
  • Safe to run multiple times (idempotent).

• Holon Linking in quint_propose:

  • New depends_on parameter to declare dependencies on other holons.
  • New decision_context parameter to group alternatives under a decision.
  • New dependency_cl parameter (1-3) for congruence level of dependencies.
  • Creates ComponentOf relations for system holons, ConstituentOf for episteme.
  • Creates MemberOf relations for decision grouping (no R propagation).
  • Added SQL indexes for efficient WLNK traversal.
  • Documented structural relations (B.1.1) in CLAUDE.md.

• Evidence Freshness Management (FPF B.3.4):

  • New waivers table for tracking temporary risk acceptance with full audit trail.
  • quint_check_decay now supports three modes:
    • Report mode (default): Shows freshness report with STALE/FRESH/WAIVED holons.
    • Deprecate mode: Downgrades hypothesis (L2→L1 or L1→L0) when evidence is terminally stale.
    • Waive mode: Records explicit risk acceptance with rationale and expiration date.
  • quint_test now accepts L2 hypotheses for evidence refresh (L2 + PASS stays L2 with fresh evidence).
  • Freshness report now shows individual evidence IDs (not just counts) for actionable output.
  • Implements WLNK principle: one expired evidence item = entire holon is STALE.
  • Natural language support: users can say "waive the benchmark until February" and the agent handles ID resolution.
  • New documentation: docs/evidence-freshness.md — practical guide to managing stale evidence.
  • Updated command documentation: q-decay.md and q3-validate.md.

• CI/CD Pipeline:

  • New GitHub Actions workflow (.github/workflows/ci.yml) for pull requests.
  • Triggers on PRs and pushes to main and dev branches.
  • Runs tests with race detector and coverage reporting.
  • Runs golangci-lint for code quality (errcheck, govet, staticcheck, unused, misspell).
  • Uses golangci-lint-action@v7 with golangci-lint v2 config schema.
  • Added .golangci.yml configuration for consistent linting.

• Pre-commit Hooks:

  • Added .pre-commit-config.yaml for use with pre-commit tool.
  • Added .githooks/pre-commit for simple git-native hooks (no dependencies).
  • Hooks include: gofmt, goimports, go build, go test, golangci-lint.
  • Setup via ./scripts/setup-hooks.sh or ./scripts/setup-hooks.sh --precommit.
  • Also checks: trailing whitespace, end-of-file, yaml syntax, large files, merge conflicts, private keys.

Changed

• Updated FPF Commands: Commands now leverage new MCP tools for computed data:

  • /q4-audit: Now calls quint_calculate_r and quint_audit_tree before recording findings.
  • /q5-decide: Now uses quint_calculate_r for final R_eff comparison before decision.
  • /q-audit: Updated to use visualization tools.
  • /q-decay: Updated to use quint_check_decay for proactive decay detection.
  • /q-status: Now proactively checks for expired evidence.

• SQLite Driver Migration: Replaced CGO-based mattn/go-sqlite3 with pure Go modernc.org/sqlite.

  • Enables CGO_ENABLED=0 builds for simplified cross-compilation.
  • Cross-compilation now works for linux/amd64, linux/arm64, darwin/*, windows/amd64.
  • Unblocks single-runner GoReleaser builds.
  • No functional changes to database behavior.

• FSM Phase Derivation: CanTransition() now uses GetPhase() (derived from DB) instead of State.Phase.

  • Phase transitions are validated against actual database state.
  • Hardens FPF enforcement against state.json manipulation.

Fixed

• Evidence Decay Bug: Evidence was stored with NULL valid_until, making /q-decay always report "no expired evidence."

  • ManageEvidence now sets a default 90-day validity period when validUntil is empty.
  • Affects all evidence added via quint_verify, quint_test, and quint_audit.

• Go Module Import Paths: Standardized import paths to use the correct module name across all packages. (PR #16, @blib)

---

v4.0.0

Added

• MCP Server Architecture:
  • This release introduces the MCP (Model Context Protocol) server as the core of Quint Code, replacing the previous prompt-only approach. The server provides structured tools for AI assistants to interact with the FPF knowledge base.
  • The MCP server has been restructured into cmd and internal packages for better organization and maintainability.
• New Commands:
  • /q-audit: Visualize assurance tree with R-scores.
  • /q-decay: Calculate epistemic debt from expired evidence.
• Command Updates:
  • /q-actualize: Reconcile the knowledge base with recent code changes. This command has been updated for better performance and accuracy.
• Renamed Commands:
  • /q2-check is now /q2-verify.
  • /q3-research and /q3-test are consolidated into /q3-validate.
  • /q1-add has been added for manually adding hypotheses.
• SQLite Database (quint.db):
  • The project now uses SQLite for deterministic FPF, ensuring consistency and reproducibility.
  • holons table now includes scope, kind, and cached_r_score columns to support advanced FPF features.
  • evidence table now includes a valid_until column for evidence decay tracking.
  • relations table now includes a congruence_level column for WLNK calculation.
• Trust & Assurance Calculator (B.3):
  • Implemented the FPF B.3 standard for calculating trust and assurance.
  • Weakest Link (WLNK): R-score is now capped by the lowest-scoring dependency in the evidence chain.
  • Congruence Penalty: Low-congruence relations between artifacts now reduce the overall reliability score.
  • Evidence Decay: Expired evidence, as determined by the valid_until field, now penalizes the R-score, introducing the concept of "epistemic debt."
  • Cycle Detection: The calculator now detects and flags circular dependencies in the evidence graph.
• Typed Reasoning (Kind-CAL):
  • Hypotheses are now classified by kind as either system (for code and architecture) or episteme (for knowledge and theory).
  • Validation logic now branches based on the hypothesis kind, allowing for more targeted and relevant analysis.
• Characteristic Space (C.16):
  • Success metrics are now defined upfront, before testing, as part of the Characteristic Space.
  • These metrics are measured during the induction phase (/q3-validate).
  • The results are recorded in the Design Rationale Record (DRR) for full traceability.
• CI/CD:
  • A new GitHub Actions workflow has been added to automate the build and release process.
• Testing:
  • Added integration tests for the assurance calculator to ensure correctness and stability.
• Error Handling:
  • Improved error handling to surface previously-silent errors with warnings.

Changed

• Project Directory: Renamed from .fpf to .quint. Migration: run /q-actualize to migrate, then delete .fpf.
• Multi-platform Support:
  • Installer now supports Claude Code, Cursor, Gemini CLI, and Codex CLI.
  • Commands are now sourced from src/commands/ and built to platform-specific formats in dist/.

Removed

• Redundant Agents: Removed legacy standalone agent files for a cleaner codebase.

---

Release: 3.2.0

Added: Process Hardening & Flexibility

Strict Phase Gating (FPF Integrity)

• Hard Block in /q1-hypothesize: Explicitly forbids generating new hypotheses if the cycle has passed Deduction. This prevents the "Helpfulness Bias" vulnerability where AI assistants might break process integrity to be "nice".
• Conditional Logic in /q2-check: The cycle phase now only advances to DEDUCTION_COMPLETE when all active L0 hypotheses are resolved. If any remain unchecked, the door stays open for extensions.

New Command: /q1-extend

• Legitimate Extension Path: A dedicated command to add a missed hypothesis during the ABDUCTION_COMPLETE phase.
• Safety Rails: Strictly blocked once DEDUCTION_COMPLETE is reached, ensuring evidence integrity (WLNK validity) during testing.

Changed

• Updated /q-status: State machine visualization now includes the (q1-extend) loop.
• Refined /q3-test & /q3-research: Reinforced checks to ensure testing only happens after deduction is fully complete.

Release 3.1.0

Accumulated release.

1. Major up to 3.0.0 due to rebranding of tool and commands (read the changelog for details)

2. Release 3.1.0

Added: Deep Reasoning Capabilities

1. Context Slicing (A.2.6)

Structured Context: .fpf/context.md is now structured into explicit slices:

• Slice: Grounding (Infrastructure, Region)
• Slice: Tech Stack (Language, Frameworks)
• Slice: Constraints (Compliance, Budget, Team)

Context-Aware Init: /q0-init now scans package.json, Dockerfile, etc., to auto-populate slices.

2. Explicit Role Injection (A.2)

Role-Swapping Prompts: Commands now enforce specific FPF roles to prevent "AI drift":

• /q1-hypothesize: ExplorerRole (Creative, Abductive)
• /q2-check: LogicianRole (Strict, Deductive)
• /q4-audit: AuditorRole (Adversarial, Normative)

Context Drift Analysis

• New Audit Step: /q4-audit now includes a mandatory Context Drift Check.
• Validation: Verifies that hypotheses generated in step 1 still match the constraints in step 4 (preventing "works on my machine" architecture).

Changed

Command Prompts: Updated q0, q1, q2, q4 to enforce the new reasoning standards.

Release: 2.2.0

Multi-Platform Support 🔥

Crucible Code now works with four AI coding tools:

• Claude Code
• Cursor
• Gemini CLI
• Codex CLI

One-Liner Install

Global (all projects)

curl -fsSL https://raw.githubusercontent.com/m0n0x41d/crucible-code/main/install.sh | bash -s -- -g

Per-project (commit to repo)

curl -fsSL https://raw.githubusercontent.com/m0n0x41d/crucible-code/main/install.sh | bash

Interactive TUI lets you pick which platforms to install.

What's New

Multi-Platform Architecture

• Adapter-based build: src/commands/ → dist/{platform}/
• Platform-specific formats (TOML for Gemini, Markdown for others)
• CI/CD ensures dist/ stays in sync

TUI Installer

• Interactive platform selection with arrow keys or vim motions (j/k)
• Global (-g) or per-project install
• Bash 3.x compatible (works on macOS out of the box)
• Uninstall support (--uninstall)

Visual Refresh

• Melted steel gradient (red → white) for ASCII banner
• New SVG banner for GitHub README

Upgrade

Just re-run the installer to get the latest commands. No breaking changes to .fpf/ structure.

Release: 2.1.0

Release v2.1.0 Summary

Here is what has been upgraded in your codebase:

1. Agentic Initialization (/fpf-0-init)

   • Now scans the repo (package.json, Dockerfile, etc.) to infer the Tech Stack.
   • Interviews the user to fill in the gaps (Scale, Budget).
   • Creates a living .fpf/context.md file that acts as the "Base Slice" for all future reasoning.

2. Strict Method/Work Distinction (/fpf-1-hypothesize)

   • The hypothesis template is now split into "1. The Method" (Plan) and "2. The Validation" (Execution).
   • This enforces FPF Pattern A.15 without forcing the user to learn the ontology.

3. NQD & Formality Tracking

   • Hypotheses now carry formality: [0-9] and novelty/complexity tags in frontmatter.
   • /fpf-status (via the logic update we discussed earlier) will calculate Min/Max Formality and Trust Index.

4. Context Awareness

   • Research and Test commands now automatically read .fpf/context.md to ensure evidence congruence is calculated against the actual project
     reality, not an abstract one.

Release: 2.0.0

Version 2.0.0 is a major release that transforms FPF from advisory guidance into an enforced reasoning framework. The core change: phases are now gates, not suggestions.

Headline Changes

1. Phase Gate Enforcement

• Commands now block invalid phase transitions. You can't /fpf-2-check without hypotheses, can't gather evidence without L1 claims.
• The ADI cycle is strictly sequential for phases 1→2→3. Audit (phase 4) can be skipped to go directly to /fpf-5-decide, but this is not recommended.

2. Transformer Mandate Enforcement

• Explicit "AWAIT HUMAN INPUT" sections at all decision points
• Claude generates options, presents analysis, then stops and waits
• No more implicit decisions buried in command output

3. Quantitative WLNK Analysis

• R_eff = R_base - Φ(CL) formula now calculated in audit
• Evidence chain tables show exactly where your argument is weakest
• Congruence penalties formalized: High=0.00, Medium=0.15, Low=0.35

4. Plausibility Filters

• Hypotheses assessed against four filters: Simplicity, Explanatory Power, Consistency, Falsifiability
• Clear verdicts: PLAUSIBLE / MARGINAL / IMPLAUSIBLE
• Ranking table for comparison

5. Better Learning Preservation

• /fpf-discard now captures key insights before cleanup
• Evidence validity windows mandatory
• Epistemic debt tracking in /fpf-decay

Breaking Changes

None. Existing .fpf/ directories work without modification.

Recommended migration:

1. Run /fpf-decay to identify evidence needing validity dates
2. Add congruence assessment to existing external evidence
3. Existing DRRs remain valid

What's New in Documentation

• README updated with phase strictness explanation and visual diagram annotations
• Commands Reference table now shows which phases are required vs optional
• "Common Mistakes to Avoid" sections in each command
• Anti-pattern tables with explanations

Upgrade Path

cd crucible-code
git pull origin main
./install.sh /path/to/your/project

Existing FPF sessions continue working. New sessions get enforced gates

Release: 1.0.0

First, the somehow stable release of crucible-code includes a basic ADI reasoning cycle from FPF, which has proven to be valuable for engineering work.