Files
selfheal-lab-apps/harness/proposals.go
T
seon f30af2a4c3 feat(ai): add AI VM services (diagnosis agent, pgvector RAG, Discord HITL bot) and harness approval endpoints
실험 설계서 6장(WP-12)의 AI VM 2층 구현.
- cmd/agent: 하네스 /diagnose 수신 → OpenAI Responses API 함수 호출 루프(읽기 전용 도구 7종:
  query_metrics, query_logs, get_probe_status, get_k8s, get_gcp_instances, get_aws_state,
  search_runbooks) → strict 스키마 submit_diagnosis(6.3) → 하네스 /proposals 등록 + HITL 게시,
  또는 신뢰도 0.5 미만·escalate 시 3층 큐 에스컬레이션. 세션·프롬프트 버전·러너북 벡터는
  pgvector(ai 스키마)에 저장. 라벨·조치 enum은 하네스 /scenarios·/actions/catalog에서 갱신.
- cmd/hitl: Discord Gateway 봇(intents GUILDS만). 승인/거부 버튼(custom_id approve:/reject:<token>),
  3초 내 DEFERRED_UPDATE_MESSAGE, 승인 시 하네스 승인 → /actions/execute → 원 메시지 편집,
  15분 무응답 시 에스컬레이션 채널로 이동. 재기동 시 미결 제안 타이머 복원.
- harness: POST /proposals/{token}/approve, /reject 추가. 제안은 진단 시점에 등록하고 승인은
  버튼 시점에 기록해 MTTA(승인-제안)가 실제 대기 시간이 되도록 함.
OpenAI·Discord 토큰은 Secret Manager 파일로만 읽고 코드·저장소에 두지 않는다.
2026-10-08 01:36:10 +09:00

113 lines
4.2 KiB
Go

package main
import (
"crypto/rand"
"encoding/hex"
"encoding/json"
"errors"
"net/http"
"time"
)
// POST /proposals: the AI VM registers a layer-2 proposal (design 6.5). Response carries the approval token that the
// Discord approval hands back; approved=true in the body records the approval directly (used by tests and by the AI VM
// after the button round-trip). Execution then goes through /actions/execute with that token.
type ProposalRequest struct {
RunID *string `json:"run_id"`
ScenarioID *string `json:"scenario_id"`
ActionID string `json:"action_id"`
Target map[string]any `json:"target"`
Diagnosis map[string]any `json:"diagnosis"`
Approved bool `json:"approved"`
ApprovedBy string `json:"approved_by"`
TTLSeconds int `json:"ttl_seconds"`
}
func (a *App) handleProposal(w http.ResponseWriter, r *http.Request) {
var req ProposalRequest
if err := json.NewDecoder(r.Body).Decode(&req); err != nil {
writeErr(w, 400, err)
return
}
if catalogEntry(req.ActionID) == nil {
writeErr(w, 400, errors.New("action_id not in the catalogue"))
return
}
if cur := a.runs.current(); cur != nil && req.RunID == nil {
req.RunID = &cur.RunID
if cur.ScenarioID != "" {
req.ScenarioID = &cur.ScenarioID
}
}
ttl := 15 * time.Minute // design 2.1: no approval within 15 minutes -> layer 3
if req.TTLSeconds > 0 {
ttl = time.Duration(req.TTLSeconds) * time.Second
}
b := make([]byte, 24)
_, _ = rand.Read(b)
token := "apr-" + hex.EncodeToString(b)
if err := a.store.InsertProposal(r.Context(), token, req.RunID, req.ScenarioID, req.ActionID, req.Target, ttl); err != nil {
writeErr(w, 500, err)
return
}
a.store.Event(r.Context(), req.RunID, req.ScenarioID, "proposal", "ai", map[string]any{"action_id": req.ActionID, "target": req.Target, "diagnosis": req.Diagnosis, "token_suffix": token[len(token)-6:]})
if req.Approved {
by := req.ApprovedBy
if by == "" {
by = "unspecified"
}
_ = a.store.ApproveProposal(r.Context(), token, by)
a.store.Event(r.Context(), req.RunID, req.ScenarioID, "approval", by, map[string]any{"action_id": req.ActionID, "token_suffix": token[len(token)-6:]})
}
writeJSON(w, 201, map[string]any{"token": token, "expires_in_s": ttl.Seconds(), "approved": req.Approved})
}
// POST /proposals/{token}/approve: the HITL bot records the human decision (design 6.5 step 4). The proposal was
// registered at diagnosis time, so MTTA = approval - proposal (design 5.4) measures the real waiting time.
func (a *App) handleProposalApprove(w http.ResponseWriter, r *http.Request) {
token := r.PathValue("token")
var req struct {
ApprovedBy string `json:"approved_by"`
}
_ = json.NewDecoder(r.Body).Decode(&req)
if req.ApprovedBy == "" {
req.ApprovedBy = "unspecified"
}
p, err := a.store.GetProposal(r.Context(), token)
if err != nil {
writeErr(w, 404, errors.New("unknown proposal"))
return
}
switch {
case p.ApprovedAt != nil:
writeErr(w, 409, errors.New("proposal already approved"))
return
case time.Now().After(p.ExpiresAt):
writeErr(w, 410, errors.New("proposal expired"))
return
}
if err := a.store.ApproveProposal(r.Context(), token, req.ApprovedBy); err != nil {
writeErr(w, 500, err)
return
}
a.store.Event(r.Context(), p.RunID, p.ScenarioID, "approval", req.ApprovedBy, map[string]any{"action_id": p.ActionID, "token_suffix": token[len(token)-6:]})
writeJSON(w, 200, map[string]any{"token": token, "approved": true, "approved_by": req.ApprovedBy})
}
// POST /proposals/{token}/reject: recorded as a note event (kind list of design 5.3); the token stays unapproved.
func (a *App) handleProposalReject(w http.ResponseWriter, r *http.Request) {
token := r.PathValue("token")
var req struct {
RejectedBy string `json:"rejected_by"`
Reason string `json:"reason"`
}
_ = json.NewDecoder(r.Body).Decode(&req)
p, err := a.store.GetProposal(r.Context(), token)
if err != nil {
writeErr(w, 404, errors.New("unknown proposal"))
return
}
a.store.Event(r.Context(), p.RunID, p.ScenarioID, "note", req.RejectedBy, map[string]any{"rejected": true, "action_id": p.ActionID, "reason": req.Reason, "token_suffix": token[len(token)-6:]})
writeJSON(w, 200, map[string]any{"token": token, "rejected": true})
}