← all publishers

Zhachory1

@zhachory1 source repo

Report this profile Claim this profile

• 16 skills
• 0 followers
• 10 hours ago last updated

1. ▌
   Backprop · zhachory1

   Backprop workflow optimizer for AI agent workflows. Use for retrospective optimization, previous execution review, workflow telemetry analysis, run history, latency/cost/failure rate, human-edit rate, rework loops, recurring failure modes, prompt/skill improvement, council regression review, shadow rollout, A/B test, promote-or-rollback changelog. Uses long-term memory MCP, code RAG, agent-fleet journal/transcripts, PR/CI data, and run telemetry.
   0
   installs
2. ▌
   Pr Review · zhachory1

   Agent PR review workflow for correctness, safety, maintainability, architecture, security, performance, tests, and fidelity to intent. A target PR URL defaults to publishing a GitHub review with the verdict-appropriate event unless the user asks for read-only, draft-only, report-only, or no posting. Produces prioritized findings, delta re-review, human merge recommendation, telemetry, and code RAG context.
   0
   installs
3. ▌
   Roadmapping · zhachory1

   Roadmap creation workflow for agents. Use for roadmap, OKR planning, project evaluation, ROI scoring, dependency mapping, capacity allocation, now/next/later, theme grouping, council review, leadership approval. Produces prioritized roadmap with value/cost/risk tradeoffs, dependencies, resource plan, approval package, telemetry, memory and code RAG context, bounded allocation/dependency loops.
   0
   installs
4. ▌
   Run Telemetry · zhachory1 bundle

   Emit structured workflow telemetry for AI agent runs. Use for run metrics, step latency, cost, tokens, model, loop count, failure rate, human-edit rate, escalations, council verdicts, validation status, PR outcomes, ML metrics, debug evidence, backprop data. Writes JSONL events to .workflow-runs/<run-id>/telemetry.jsonl or caller-provided store for later optimization.
   0
   installs
5. ▌
   Ml Experiments · zhachory1

   ML experiment workflow from problem formulation through reproducible handoff. Use for model experiments, metric definition, council signoff, data pipeline, EDA, tracking/logging, hypothesis generation, baseline, training, validation, artifacts. Enforces locked metrics, leakage checks, experiment tracking, hypothesis/train/validate loop, compute caps, tried-rejected ledger, handoff, telemetry, long-term memory, and code RAG.
   0
   installs
6. ▌
   Plan To Launch · zhachory1

   End-to-end software development workflow for AI agents. Use for plan-to-launch, dev workflow, PRD, DD, design doc, council on docs, implementation plan, ship, code-reviewer, council on review and code, launch PR, merge readiness. Enforces thought-out plan before implementation, bounded PRD/DD council loop, ship/review loop, design kickback, human approval, agent-fleet awareness, telemetry, long-term memory, and code RAG.
   0
   installs
7. ▌
   Grounding Brief · zhachory1

   Universal context-gathering skill for AI agent workflows. Use for grounding brief, gather context, step zero, linked docs, tickets, code search, logs, traces, metrics, prior runs, memory MCP, gbrain, code RAG, repo-index, agent-fleet journal, PR context, investigation evidence, roadmap inputs, ML experiment setup. Produces a standard brief with sources, assumptions, open questions, risks, and confidence.
   0
   installs
8. ▌
   Project Ideation · zhachory1

   Project ideation workflow for OKR planning and feature discovery. Use for project ideation, feature ideas, goal deconstruction, gap analysis, customer/data-backed unmet needs, cross-functional jam, multi-persona council, project framing, success metric, technical feasibility, engineering lift. Produces framed candidates ready for roadmap evaluation with telemetry, memory, code RAG, and bounded diverge/converge loops.
   0
   installs
9. ▌
   Adversarial Probe · zhachory1

   Exploratory adversarial testing workflow for agents. Use for stress testing, load/soak/spike test, chaos engineering, fault injection, fuzzing, boundary/malformed input, resource exhaustion, concurrency/race probing, red-teaming, jailbreak, prompt injection, tool-abuse, spec-gaming, eval robustness, agent safety-boundary probing, find where it breaks, attack surface mapping. Targets software services, software tools, and AI-agent workflows. Enforces locked breakage thresholds before attack, threat model, bounded attack loop, mandatory approval before destructive/live probes, confirmed-breakage handoff to debug-investigation, telemetry, long-term memory, and code RAG.
   0
   installs
10. ▌
    Handoff Packaging · zhachory1

    Package workflow outputs for the next agent or human. Use for launch PR package, ML artifacts handoff, debug RCA handoff, roadmap approval packet, project ideation handoff, backprop changelog, implementation handoff. Produces summary, evidence, decisions, files, artifacts, metrics, known limitations, next steps, and telemetry.
    0
    installs
11. ▌
    Hypothesis Testing · zhachory1

    Reusable hypothesis loop for ML experiments and debugging. Use for hypothesis generation, confirmation/refutation, root cause analysis, train/validate cycles, evidence tests, leakage checks, tried/rejected ledger, experiment iteration, debugging traces. Produces hypotheses with evidence, tests, verdicts, ruled-out alternatives, loop stop conditions, and telemetry.
    0
    installs
12. ▌
    Debug Investigation · zhachory1

    Debug and investigation workflow for agents. Use for production issue, failing test, API bug, performance regression, data issue, incident, root cause analysis, evidence gathering, logs, traces, metrics, reproduction, codebase tracing, hypothesis confirmation, RCA handoff. Enforces symptom definition, bounded hypothesis loop, evidence ledger, human escalation, telemetry, long-term memory, code RAG, and prevention actions.
    0
    installs
13. ▌
    Human Approval Gate · zhachory1

    Prepare decision package for qualified human signoff. Use for PR launch, merge/no-merge, roadmap approval, ML deploy handoff, metric changes, council blockers, destructive actions, production risk, budget cap escalation. Produces concise yes/no ask, recommendation, evidence, risks, alternatives, rollback, and telemetry; never auto-approves accountable decisions.
    0
    installs
14. ▌
    Structured Doc Authoring · zhachory1 bundle

    Author and revise structured workflow docs for agents: PRD, DD, design doc, implementation plan, RCA, ADR, experiment plan, roadmap proposal, handoff doc. Use for doc writing, doc quality gate, templates, acceptance criteria, risks, tradeoffs, decision records, prevention actions, Mermaid diagrams, Graphviz/DOT diagrams, visual explanations. Produces concise docs with fixed sections, visuals, source links, open questions, and telemetry.
    0
    installs
15. ▌
    Success Criteria Metrics · zhachory1

    Define success criteria and metrics before work starts. Use for PRD outcomes, ML offline/online metrics, experiment yardsticks, roadmap ROI, ideation framing, debug done criteria, backprop workflow optimization, launch gates, acceptance criteria. Produces measurable targets, guardrails, measurement plan, anti-metric-shopping lock, baselines, decision thresholds, and telemetry.
    0
    installs
16. ▌
    Task Decomposition Planning · zhachory1

    Turn approved goals, PRDs, DDs, ML experiment plans, roadmaps, or investigations into ordered dependency-aware tasks. Use for implementation plan writing, experiment plan, roadmap sequencing, acceptance criteria, blockers, atomic PRs, verifiable tasks, handoff to ship agents. Produces tasks with owners, dependencies, validation, acceptance criteria, risk, and telemetry.
    0
    installs