observable-gepa-migration
Migrate DSPy GEPA usage in DSPy from the original to gepa-observable. This makes it possible for teams to clearly review each iteration and the lineage to understand how their prompt is evolving. The repository offers a web dashboard for monitoring, but requires a custom GEPA fork that provides custom observers and LM call logging. Use when developers want to add observability to GEPA optimization.
What this skill does
# Migrate DSPy GEPA to gepa-observable
Migrate existing DSPy GEPA code to gepa-observable for integrated observability:
real-time dashboard, custom observer callbacks, and LM call capture working together.
## Quick Migration
### Step 1: Install gepa-observable
```bash
pip install dspy-gepa-logger
```
### Step 2: Change Import
```python
# Before
from dspy.teleprompt import GEPA
# After
from gepa_observable import GEPA
```
### Step 3: Add Observability (Full System)
```python
from gepa_observable import GEPA
optimizer = GEPA(
metric=my_metric,
auto="medium",
# Observability system:
server_url="http://your-server:3000", # Dashboard integration
project_name="My Project",
capture_lm_calls=True, # LM call logging
capture_stdout=True, # Console capture
verbose=True, # LoggingObserver
observers=[MyCustomObserver()], # Custom callbacks
)
```
### Step 4: Everything Else Stays the Same
- Metric functions (5-arg signature already required by DSPy GEPA)
- Program definitions (ChainOfThought, ReAct, etc.)
- Training/validation data format
- Budget params (auto, max_full_evals, max_metric_calls)
- compile() call: `optimizer.compile(student=program, trainset=train, valset=val)`
## The Observability System
gepa-observable provides an **integrated observability system** where all components work together:
| Component | Purpose | Enabled By |
|-----------|---------|------------|
| **ServerObserver** | Sends events to web dashboard | `server_url` param |
| **LoggingObserver** | Console output with summaries | `verbose=True` |
| **LM Call Logger** | Captures all LM invocations | `capture_lm_calls=True` |
| **Custom Observers** | Your callbacks for any event | `observers=[...]` |
All observers receive the **same 8 lifecycle events**:
1. `SeedValidationEvent` - Initial validation scores
2. `IterationStartEvent` - Each optimization iteration
3. `MiniBatchEvalEvent` - Minibatch evaluations
4. `ReflectionEvent` - Proposed prompt changes
5. `AcceptanceDecisionEvent` - Accept/reject decisions
6. `ValsetEvalEvent` - Full validation evaluations
7. `MergeEvent` - Candidate merge operations
8. `OptimizationCompleteEvent` - Final results
## Custom Observer Example
```python
class MyObserver:
def on_seed_validation(self, event):
avg = sum(event.valset_scores.values()) / len(event.valset_scores)
print(f"Seed score: {avg:.2%}")
def on_iteration_start(self, event):
print(f"Iteration {event.iteration}, parent score: {event.parent_score:.2%}")
def on_reflection(self, event):
for comp, text in event.proposed_texts.items():
print(f"Proposing for {comp}: {text[:100]}...")
def on_valset_eval(self, event):
if event.is_new_best:
print(f"NEW BEST: {event.valset_score:.2%}")
def on_optimization_complete(self, event):
print(f"Done! Best: {event.best_score:.2%} in {event.total_iterations} iters")
# Use with other observers - they all work together
optimizer = GEPA(
metric=my_metric,
auto="medium",
server_url="http://localhost:3000",
observers=[MyObserver()],
verbose=True, # Also keep LoggingObserver
)
```
## Migration Checklist
### For Notebooks
1. [ ] Install: `pip install dspy-gepa-logger`
2. [ ] Change import: `from gepa_observable import GEPA`
3. [ ] Add `server_url` for dashboard
4. [ ] Add custom observers if needed
5. [ ] Run cells in order - API is 100% compatible
### For Scripts
1. [ ] Install: `pip install dspy-gepa-logger`
2. [ ] Find all `from dspy.teleprompt import GEPA`
3. [ ] Replace import
4. [ ] Add observability parameters to GEPA constructor
5. [ ] No other changes needed
## References
- `references/api-reference.md` - Complete parameter docs, observer protocol, all event types
- `references/examples.md` - Full before/after code examples
Related in AI Agents
skill-development
IncludedComprehensive meta-skill for creating, managing, validating, auditing, and distributing Claude Code skills and slash commands (unified in v2.1.3+). Provides skill templates, creation workflows, validation patterns, audit checklists, naming conventions, YAML frontmatter guidance, progressive disclosure examples, and best practices lookup. Use when creating new skills, validating existing skills, auditing skill quality, understanding skill architecture, needing skill templates, learning about YAML frontmatter requirements, progressive disclosure patterns, tool restrictions (allowed-tools), skill composition, skill naming conventions, troubleshooting skill activation issues, creating custom slash commands, configuring command frontmatter, using command arguments ($ARGUMENTS, $1, $2), bash execution in commands, file references in commands, command namespacing, plugin commands, MCP slash commands, Skill tool configuration, or deciding between skills vs slash commands. Delegates to docs-management skill for official documentation.
reprompter
IncludedTransform messy prompts into well-structured, effective prompts — single or multi-agent. Use when: "reprompt", "reprompt this", "clean up this prompt", "structure my prompt", rough text needing XML tags and best practices, "reprompter teams", "repromptception", "run with quality", "smart run", "smart agents", multi-agent tasks, audits, parallel work, anything going to agent teams. Don't use when: simple Q&A, pure chat, immediate execution-only tasks. See "Don't Use When" section for details. Outputs: Structured XML/Markdown prompt, quality score (before/after), optional team brief + per-agent sub-prompts, agent team output files. Success criteria: Single mode quality score ≥ 7/10; Repromptception per-agent prompt quality score 8+/10; all required sections present, actionable and specific.
adaptive-compaction
IncludedAdaptive add-on policy and recovery layer that decides WHEN to compact, prune, snapshot, or fork -- replacing fixed-percent auto-compaction across Claude Code, Codex, and MCP-capable hosts. Trigger on auto-compact timing or damage: "when should I compact", "is it safe to compact now or start a fresh session", "auto-compact fires too early/mid-task", "switching to an unrelated task but the window still has space", "context rot", "answers get worse the longer the session runs", "the agent forgot the plan or my decisions after it summarized", "add a layer on top that manages context without changing the agent", raising autoCompactWindow to give the policy room, or installing/tuning a cross-tool compaction policy or PreCompact hook -- even when "compaction" is never said but the problem is context-window pressure or post-summarization memory loss. Do NOT use to summarize a conversation, build RAG, write a summarization prompt (decides WHEN not HOW), or answer max-context-length trivia.
agent-skill-creator
IncludedCreate cross-platform agent skills from workflow descriptions. Activates when users ask to create an agent, automate a repetitive workflow, create a custom skill, or need advanced agent creation. Triggers on phrases like create agent for, automate workflow, create skill for, every day I have to, daily I need to, turn process into agent, need to automate, create a cross-platform skill, validate this skill, export this skill, migrate this skill. Supports single skills, multi-agent suites, transcript processing, template-based creation, interactive configuration, cross-platform export, and spec validation.
llm-wiki
IncludedUse when building or maintaining a persistent personal knowledge base (second brain) in Obsidian where an LLM incrementally ingests sources, updates entity/concept pages, maintains cross-references, and keeps a synthesis current. Triggers include "second brain", "Obsidian wiki", "personal knowledge management", "ingest this paper/article/book", "build a research wiki", "compound knowledge", "Memex", or whenever the user wants knowledge to accumulate across sessions instead of being re-derived by RAG on every query.
skill-master
IncludedAgent Skills authoring, evaluation, and optimization. Create, edit, validate, benchmark, and improve skills following the agentskills.io specification. Use when designing SKILL.md files, structuring skill folders (references, scripts, assets), ingesting external documentation into skills, running trigger evals, benchmarking skill quality, optimizing descriptions, or performing blind A/B comparisons. Keywords: agentskills.io, SKILL.md, skill authoring, eval, benchmark, trigger optimization.