Pandoscope: a nice decision-driven orchestrator, self-catalyzing, -observing, -pruning & -explaining
Pandoscope publishes agent skills for test-driven development with a red-green-refactor commit protocol, scored multiple-choice grilling of plans and decisions, fresh-session probes for testing session machinery, and architecture decision records in docs/adr/. Its skills also cover placing DECISION markers in agent-generated code, preparing sessions for context compaction, and terse chat output. Much of the collection concerns keeping agent work auditable, with decisions recorded and tests written before implementation.
Skills
14 skills · 1 stars · MIT
Caveman Terse chat that stays easy to parse.
Documenting Decisions Routes human attention to decisions that matter in agent-generated code.
Domain Modeling Build and sharpen a project's domain model.
Grilling Grill the user about a plan, decision, or idea via scored multiple-choice questions.
Handing Off Prepare a session for compaction: write the handoff, mark it fresh, propose /compact focus lines.
Run Probe Run a fresh-session probe against this environment from a checked-in probe reference.
Take Probe Execute a checked-in probe fixture in this session.
Tdd Test-driven development via red-green-refactor with an auditable commit protocol.
Thread Ledger Records a session's open work as append-only events in the session-memory store, and renders the…
To Spec Turn the current conversation into a spec published to the project tracker — no interview, just s…
To Tickets Break a plan, spec, or the current conversation into tracer-bullet tickets with blocking edges, p…
Writing Adrs Format, numbering, and when-to-write criteria for Architecture Decision Records in docs/adr/.
Writing Prose Prose rules per surface, and the loop that writes, rewrites and reviews prose against them.
Writing Skills Write and edit agent skills that take the same process every run.