#961 by GitHub stars
SynaLinks/synalinks-skills
Use for anything involving the Synalinks neuro-symbolic LM framework (Keras-inspired): DataModel/Field/Input, JSON operators (+ & | ^ ~), synalinks.ops; LanguageModel/EmbeddingModel and provider prefixes; DecisionModel (TypeSafe jev decision models, decision_model=, min_confidence/threshold); the Pr
History
Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.
View weekly changes
skills/synalinks/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Network or broad shell → After Network or broad shell
092970a578e5 → 092970a578e5
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Network or broad shell
First observation → 092970a578e5
No earlier version was observed for comparison.
908GitHub stars
0stars in 7 days
1skill in this repository
0commits in 30 days
Heat
0.4- Star growth, last 7 daysAhead of 6% of tracked repositories
- Star growth, last 24 hoursAhead of 3% of tracked repositories
- Rank movement, last 7 daysAhead of 4% of tracked repositories
- Commits, last 30 daysAhead of 0% of tracked repositories
- Accounts asking about it through MCP, last 7 daysNot available
- Hot votes this weekAhead of 0% of tracked repositories
Heat never changes the order by stars, and sponsorship never touches it. How heat works
Stars over time
Individual skills in this repository
- synalinks
skills/synalinks
How it works
synalinks
When your agent uses it
Use for anything involving the Synalinks neuro-symbolic LM framework (Keras-inspired): DataModel/Field/Input, JSON operators (+ & | ^ ~), synalinks.ops; LanguageModel/EmbeddingModel and provider prefixes; DecisionModel (TypeSafe jev decision models, decision_model=, min_confidence/threshold); the Program class and its four APIs, save/load; modules (Generator, ChainOfThought, SelfCritique, PythonSynthesis, custom Module); control flow (Decision, MultiDecision, Branch, And/Or/Xor, guards); agents (FunctionCallingAgent, RLM, DeepAgent, Tool, MCP); KnowledgeBase/RAG; training (compile/fit/evaluate/predict, callbacks, KerasTuner tuners); rewards (ExactMatch, CosineSimilarity, LMAsJudge, ProgramAsJudge, RubricsAsJudge and presets like Faithfulness/Toxicity, AgentAsJudge, DeepAgentAsJudge, RLMAsJudge, ComposableReward, BatchReward) and metrics; optimizers (RandomFewShot, OMEGA); datasets, visualization. Synalinks is Keras-shaped: without guidance LMs mix Keras/LangChain/DSPy syntax.
What it loads into context
Frontmatter
255 tokensAlways available to the agent
SKILL.md
5,053 tokensWhen the skill is activated
Reference documents
32,481 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Core abstractions
- Universal gotchas (apply everywhere)
- Quick Start
- Core: DataModel, operators, models
- Programs: the four building APIs
- Modules: generation & custom subclassing
- Control flow: routing, parallelism, guards
- Agents: tools, RLM, DeepAgent, MCP
- Knowledge & RAG
- Training (in-context RL)
- Decision models
- Rewards & metrics
- Optimizers
- Providers
- Datasets & visualization
- Scripts → logs index
- Reference docs
Scripts executed separately
scripts/chain_of_thought.pyscripts/conditional_branches.pyscripts/custom_dataset.pyscripts/custom_module.pyscripts/custom_reward.pyscripts/decision_model.pyscripts/deep_agent.pyscripts/four_apis.pyscripts/guard_patterns.pyscripts/lmstudio_setup.pyscripts/omega_example.pyscripts/rag_example.pyscripts/rlm_agent.pyscripts/simple_qa.pyscripts/subclassing_modules.pyscripts/tool_agent.pyscripts/training_example.pyscripts/visualization.py
Files
SKILL.mdcore19,456 bytesSKILL.mdcore19,456 bytesreferences/agents-tools.mddoc16,479 bytesreferences/api-reference.mddoc12,584 bytesreferences/chain_of_thought.logasset8,732 bytesreferences/conditional_branches.logasset33,051 bytesreferences/control-flow.mddoc10,199 bytesreferences/custom_dataset.logasset477 bytesreferences/custom_module.logasset36,856 bytesreferences/custom_reward.logasset10,005 bytesreferences/data-models.mddoc6,985 bytesreferences/decision-models.mddoc9,632 bytesreferences/deep_agent.logasset112,142 bytesreferences/four_apis.logasset13,747 bytesreferences/guard_patterns.logasset54,751 bytesreferences/knowledge-base.mddoc10,105 bytesreferences/lmstudio_setup.logasset24 bytesreferences/modules-catalog.mddoc14,714 bytesreferences/omega_example.logasset1,572 bytesreferences/optimizers.mddoc6,984 bytesreferences/programs.mddoc10,421 bytesreferences/rag_example.logasset281,844 bytesreferences/rewards-metrics.mddoc16,629 bytesreferences/rlm_agent.logasset53,560 bytesreferences/simple_qa.logasset3,136 bytesreferences/subclassing_modules.logasset3,448 bytesreferences/tool_agent.logasset117,971 bytesreferences/training-guide.mddoc18,229 bytesreferences/training_example.logasset2,184 bytesreferences/visualization.logasset2,650 bytesscripts/chain_of_thought.pyruns1,500 bytesscripts/conditional_branches.pyruns2,493 bytesscripts/custom_dataset.pyruns4,400 bytesscripts/custom_module.pyruns5,084 bytesscripts/custom_reward.pyruns2,533 bytesscripts/decision_model.pyruns4,173 bytesscripts/deep_agent.pyruns2,332 bytesscripts/four_apis.pyruns5,493 bytesscripts/guard_patterns.pyruns6,775 bytesscripts/lmstudio_setup.pyruns3,168 bytesscripts/omega_example.pyruns3,812 bytesscripts/rag_example.pyruns2,577 bytesscripts/rlm_agent.pyruns1,784 bytesscripts/simple_qa.pyruns1,509 bytesscripts/subclassing_modules.pyruns5,470 bytesscripts/tool_agent.pyruns4,089 bytesscripts/training_example.pyruns3,877 bytesscripts/visualization.pyruns2,499 bytes
Read SKILL.md
# Synalinks
Synalinks is an open-source, Keras-inspired framework for building neuro-symbolic
LM applications (RAGs, agents, self-evolving reasoning) with in-context
reinforcement learning. This skill is the single entry point: each section below
is a scannable overview that points to a **runnable script + its captured run log**
and a **deep-dive reference doc** under `references/`.
## Core abstractions
- **DataModel** — Pydantic-style schema for structured I/O (replaces tensors).
- **Module** — async computational unit processing JSON data (replaces layers).
- **Program** — DAG of modules with conditional logic (replaces models).
- **Reward / Optimizer** — guide training by *maximizing reward* (no gradients;Skills included
- synalinks
Repository facts
- License
- Apache-2.0
- Created
- Jun 17, 2023
- Last pushed
- Sep 28, 2026
- Forks
- 83
- Open issues
- 1
- Project activity
- Active
Rank over the last 12 weeks
Now #961 · best #693 · down 268 in 30 days
- #943 on Oct 31
- #945 on Nov 1
- #952 on Nov 2
- #957 on Nov 3
- #958 from Nov 4 to Nov 5
- #961 since Nov 6
9 earlier periods
- #640 on Aug 1
- #576 on Aug 7
- #615 on Aug 13
- #693 on Sep 16
- #738 on Oct 18
- #915 on Oct 25
- #931 on Oct 28
- #939 on Oct 29
- #941 on Oct 30
Before you install
Network or broad shellAutomated signals, not a guarantee
Checked Nov 4, 2026
Not fully analyzed: some files were too long, unreadable or missing. This never lowers the level.
Show 2 notes
skills/synalinks/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedskills/synalinks/references/rag_example.log:10109Analysis incomplete: regex work limit reached
- No broad tool permissions
- 5 dangerous shell commandsin 2 files: synalinks/references/agents-tools.md:25 synalinks/scripts/tool_agent.py:53
- 24 network referencesin 4 files: synalinks/scripts/lmstudio_setup.py:17 synalinks/references/agents-tools.md:204 synalinks/references/training-guide.md:427 and 1 more
- No hidden or obfuscated text
- No instruction overrides
- 18 executable scriptsin 18 files: synalinks/scripts/chain_of_thought.py:1 synalinks/scripts/conditional_branches.py:1 synalinks/scripts/custom_dataset.py:1 and 15 more
- 15 references to secretsin 5 files: synalinks/SKILL.md:221 synalinks/scripts/lmstudio_setup.py:5 synalinks/references/decision-models.md:16 and 2 more
- No requests to vote in Hot
Show all 62 signals by file
skills/synalinks/SKILL.md
- Access to secretsline 221
[Credential-related source omitted] - Access to secretsline 237
[Credential-related source omitted] - Network accessline 283
Some prefixes are rewritten in `__init__`: `ollama/`→`ollama_chat/` (default `api_base` `http://localhost:11434`), `vllm/`→`hosted_vllm/` (`HOSTED_VLLM_API_BASE` - Network accessline 284
`api_base` `http://localhost:11434`), `vllm/`→`hosted_vllm/` (`HOSTED_VLLM_API_BASE` - Network accessline 285
or `http://localhost:8000`), `doubleword/`→`openai/` + the Doubleword `api_base`. - Access to secretsline 289
[Credential-related source omitted] - Access to secretsline 291
[Credential-related source omitted] - Access to secretsline 334
[Credential-related source omitted]
skills/synalinks/references/agents-tools.md
- Dangerous shell operationline 25
result = eval(expression, {"__builtins__": None}, {}) - Dangerous shell operationline 94
result = eval(expression, {"__builtins__": None}, {}) - Network accessline 204
mcp_client = synalinks.MultiServerMCPClient({ "math_server": { "url": "http://localhost:8183/mcp/", - Network accessline 205
"math_server": { "url": "http://localhost:8183/mcp/", "transport": "streamable_http", - Network accessline 206
"url": "http://localhost:8183/mcp/", - Network accessline 208
}, "search_server": { "url": "http://localhost:8184/mcp/", - Network accessline 209
"search_server": { "url": "http://localhost:8184/mcp/", "transport": "streamable_http", - Network accessline 210
"url": "http://localhost:8184/mcp/", - Dangerous shell operationline 244
return {"result": eval(expression, {"__builtins__": None}, {})} - Network accessline 330
> Paper: [Recursive Language Models](https://arxiv.org/abs/2512.24601). - Dangerous shell operationline 446
return {"result": eval(expression, {"__builtins__": None}, {}), "log": "Success"}
skills/synalinks/references/api-reference.md
- Access to secretsline 292
[Credential-related source omitted]
skills/synalinks/references/decision-models.md
skills/synalinks/references/training-guide.md
- Network accessline 427
OMEGA is based on: - **Dominated Novelty Search**: [arxiv.org/html/2502.00593v1](https://arxiv.org/html/2502.00593v1) - Network accessline 428
- **Dominated Novelty Search**: [arxiv.org/html/2502.00593v1](https://arxiv.org/html/2502.00593v1) - Network accessline 429
- **GEPA** (prompt evolution): [arxiv.org/pdf/2507.19457](https://arxiv.org/pdf/2507.19457) - Network accessline 430
- **Chain-of-Thought**: [arxiv.org/abs/2201.11903](https://arxiv.org/abs/2201.11903)
skills/synalinks/scripts/chain_of_thought.py
- Executable scriptline 1
#!/usr/bin/env python3
skills/synalinks/scripts/conditional_branches.py
- Executable scriptline 1
#!/usr/bin/env python3
skills/synalinks/scripts/custom_dataset.py
- Executable scriptline 1
#!/usr/bin/env python3
skills/synalinks/scripts/custom_module.py
- Executable scriptline 1
#!/usr/bin/env python3
skills/synalinks/scripts/custom_reward.py
- Executable scriptline 1
#!/usr/bin/env python3
skills/synalinks/scripts/decision_model.py
skills/synalinks/scripts/deep_agent.py
- Executable scriptline 1
#!/usr/bin/env python3
skills/synalinks/scripts/four_apis.py
- Executable scriptline 1
#!/usr/bin/env python3
skills/synalinks/scripts/guard_patterns.py
- Executable scriptline 1
#!/usr/bin/env python3
skills/synalinks/scripts/lmstudio_setup.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 5
[Credential-related source omitted] - Network accessline 17
The __main__ demo points the same helper at a *live* local OpenAI-compatible server — ollama at http://localhost:11434/v1 — so it runs without LMStudio. - Network accessline 18
server — ollama at http://localhost:11434/v1 — so it runs without LMStudio. - Network accessline 29
def create_lmstudio_language_model( model_name: str, api_base: str = "http://localhost:1234/v1", - Network accessline 30
model_name: str, api_base: str = "http://localhost:1234/v1", max_tokens: int = 4096, - Network accessline 31
api_base: str = "http://localhost:1234/v1", - Access to secretsline 33
[Credential-related source omitted] - Access to secretsline 42
[Credential-related source omitted] - Access to secretsline 45
[Credential-related source omitted] - Network accessline 66
def create_vllm_language_model( model_name: str, api_base: str = "http://localhost:8000/v1", - Network accessline 67
model_name: str, api_base: str = "http://localhost:8000/v1", **kwargs, - Network accessline 68
api_base: str = "http://localhost:8000/v1", - Network accessline 82
lm = create_lmstudio_language_model( "mistral", api_base="http://localhost:11434/v1" - Network accessline 83
"mistral", api_base="http://localhost:11434/v1"
skills/synalinks/scripts/omega_example.py
- Executable scriptline 1
#!/usr/bin/env python3
skills/synalinks/scripts/rag_example.py
- Executable scriptline 1
#!/usr/bin/env python3
skills/synalinks/scripts/rlm_agent.py
- Executable scriptline 1
#!/usr/bin/env python3
skills/synalinks/scripts/simple_qa.py
- Executable scriptline 1
#!/usr/bin/env python3
skills/synalinks/scripts/subclassing_modules.py
- Executable scriptline 1
#!/usr/bin/env python3
skills/synalinks/scripts/tool_agent.py
skills/synalinks/scripts/training_example.py
- Executable scriptline 1
#!/usr/bin/env python3
skills/synalinks/scripts/visualization.py
- Executable scriptline 1
#!/usr/bin/env python3
How the signals fit together
Facts from a static read of the files at the pinned commit. They describe what the code does, not why.
Some files could not be read; the facts below still hold.
Signal confidence: 1 fact
- Access to secrets in prose of the SKILL.mdsynalinks/SKILL.md:237Low confidence
Works with
| Agent | Status | Evidence | Install to |
|---|---|---|---|
| Claude Code | Compatible | Standard SKILL.md skills/synalinks/SKILL.md | .claude/skills/<skill-name>/ |
| Codex | Compatible | Standard SKILL.md skills/synalinks/SKILL.md | .agents/skills/<skill-name>/ |
| Cursor | Compatible | Standard SKILL.md skills/synalinks/SKILL.md | .cursor/skills/<skill-name>/ |
| Gemini CLI | Compatible | Standard SKILL.md skills/synalinks/SKILL.md | .gemini/skills/<skill-name>/ |
| GitHub Copilot | Compatible | Standard SKILL.md skills/synalinks/SKILL.md | .github/skills/<skill-name>/ |
| OpenCode | Compatible | Standard SKILL.md skills/synalinks/SKILL.md | .opencode/skills/<skill-name>/ |
- Native The repository ships a manifest, folder or frontmatter field for this agent.
- Compatible A standard SKILL.md that this agent can read.
- Unknown No evidence either way.
Derived automatically from paths, manifests and frontmatter. Not tested on each agent.
Explore alternatives and comparisons
Add this badge to your README
It updates every night and links to this page.
[](https://hotskills.dev/skill/synalinks-synalinks-skills/)