What is the continuity kernel and what does it do?

It injects bounded continuity context into llm_input, persists Soul Card + Mission Ticket state with deterministic runtime contracts, and supports fail-open operation along with shadow-mode eval receipts.

Where are runtime contracts stored, and can I change the path?

Default storage is a SQLite database at ~/.local/state/continuity-kernel/continuity.db; you can override it with the CONTINUITY_KERNEL_DB_PATH environment variable.

How do I run shadow-mode evaluations and obtain receipts?

Use run_shadow_eval.py with appropriate layer, chunk, and suite settings; results can be appended to an artifacts file for auditing and review, and you can supply traces or synthetic data if needed.

continuity-kernel

npx machina-cli add skill cacheforge-ai/cacheforge-skills/continuity-kernel --openclaw

Files (1)

SKILL.md

3.7 KB

When to use this skill

Use this skill when the user wants to:

Inject bounded continuity context into llm_input without manual configuration
Persist Soul Card + Mission Ticket state with deterministic runtime contracts (durable default SQLite path, zero-config)
Keep agent behavior fail-open if continuity logic or storage has an error
Generate reproducible runtime proof receipts for audits and reviews
Run shadow-mode evals and append chunk-level quality snapshots

Commands

# Run full test suite
python3 -m unittest discover -s tests/continuity-kernel -p 'test_*.py'

# Generate deterministic runtime contract proof artifact
python3 skills/continuity-kernel/generate_runtime_contract_proof.py

# SC-02 core-shadow gate (trace optional, synthetic allowed for this suite)
python3 skills/continuity-kernel/run_shadow_eval.py --layer 'Soul Card' --chunk 'SC-02' --suite core-shadow --runs 100 --append artifacts/continuity-kernel/p0-evals.json

# SC-02 perturb robustness gate (trace-backed, deterministic selector + size_only)
python3 skills/continuity-kernel/run_shadow_eval.py --layer 'Soul Card' --chunk 'SC-02' --suite memoryarena-mini-perturb --perturb-profile deletion --selector-mode deterministic --compaction-policy size_only --runs 100 --trace-jsonl tests/continuity-kernel/fixtures/shadow_eval_trace_memoryarena_mini_perturb.jsonl --append artifacts/continuity-kernel/p0-evals.json

# SC-02 perturb robustness gate (trace-backed, dual-route + attention-preserving)
python3 skills/continuity-kernel/run_shadow_eval.py --layer 'Soul Card' --chunk 'SC-02' --suite memoryarena-mini-perturb --perturb-profile deletion --selector-mode dual_route_experimental --compaction-policy attention_preserving_experimental --runs 100 --trace-jsonl tests/continuity-kernel/fixtures/shadow_eval_trace_memoryarena_mini_perturb.jsonl --append artifacts/continuity-kernel/p0-evals.json

Options

run_shadow_eval.py options:

--layer TEXT — Active P0 layer name
--chunk TEXT — Active chunk id/name
--suite core-shadow|memoryarena-mini|memoryarena-mini-perturb — Task suite for eval scoring
--perturb-profile none|deletion|noise_injection|reorder — Deterministic perturb profile for memoryarena-mini-perturb
--selector-mode deterministic|dual_route_experimental — Selector mode tag for A/B eval receipts
--compaction-policy size_only|attention_preserving_experimental — Compaction policy label for dropped-field A/B eval comparison
--runs N — Number of evaluated shadow runs (default: 100)
--trace-jsonl PATH — JSONL file containing observed shadow outcomes (repeatable)
--trace-dir PATH — Directory scanned for *.jsonl observed shadow outcomes (repeatable)
--append PATH — JSON file path to append snapshot results
--allow-synthetic — Permit synthetic fallback for non-core suites when traces are unavailable (forced pass=false)
--artifacts-root PATH — Directory for summary receipts (default: ~/.cache/continuity-kernel/p0-evals)

Runtime persistence defaults:

ContinuityHookAdapter() uses ~/.local/state/continuity-kernel/continuity.db
Override default path with CONTINUITY_KERNEL_DB_PATH
Use :memory: only as explicit opt-in test mode

Source

git clone https://github.com/cacheforge-ai/cacheforge-skills/blob/main/skills/continuity-kernel/SKILL.mdView on GitHub

Overview

This skill injects bounded continuity context into llm_input and persists Soul Card + Mission Ticket state using deterministic runtime contracts (default SQLite). It ensures fail-open behavior on continuity or storage errors and can generate reproducible receipts for audits, plus support shadow-mode evals with chunk-level snapshots.

How This Skill Works

Technically, a ContinuityHookAdapter injects bounded continuity context into llm_input and persists Soul Card + Mission Ticket state using deterministic runtime contracts stored in SQLite. The default path is ~/.local/state/continuity-kernel/continuity.db, overrideable via CONTINUITY_KERNEL_DB_PATH, with :memory: allowed only for explicit test mode. Shadow-mode evals generate appendable runtime receipts and chunk-level quality snapshots for audits.

When to Use It

Inject bounded continuity context into llm_input without manual configuration
Persist Soul Card + Mission Ticket state with deterministic runtime contracts (durable default SQLite path, zero-config)
Keep agent behavior fail-open if continuity logic or storage has an error
Generate reproducible runtime proof receipts for audits and reviews
Run shadow-mode evals and append chunk-level quality snapshots

Quick Start

Step 1: Run full test suite: python3 -m unittest discover -s tests/continuity-kernel -p 'test_*.py'
Step 2: Generate deterministic runtime contract proof: python3 skills/continuity-kernel/generate_runtime_contract_proof.py
Step 3: Run a shadow eval: python3 skills/continuity-kernel/run_shadow_eval.py --layer 'Soul Card' --chunk 'SC-02' --suite core-shadow --runs 100 --append artifacts/continuity-kernel/p0-evals.json

Best Practices

Limit injected continuity context to avoid bloated prompts and latency
Rely on the default SQLite persistence or set CONTINUITY_KERNEL_DB_PATH for durable storage
Ensure continuity logic errors are non-fatal to preserve fail-open behavior
Run shadow-mode evaluations with run_shadow_eval to generate chunk-level receipts
Supply trace data or enable synthetic traces only when traces are unavailable (--allow-synthetic)

Example Use Cases

Inject bounded continuity context into an ongoing chat to maintain consistent Soul Card behavior across turns
Persist Soul Card + Mission Ticket state across restarts using the default SQLite-backed path
Generate reproducible runtime proof receipts for an internal audit
Run SC-02 core-shadow eval with 100 runs and append to artifacts/p0-evals.json
Override the DB path with CONTINUITY_KERNEL_DB_PATH to isolate environments or test modes

Frequently Asked Questions

Add this skill to your agents