nullkey
6d8c8650cd
Two intertwined agent safety nets that share the same files:
1. --dry-run flag for offline preview of mutation commands
2. Risk: metadata + SetRisk helper for destructive command surfaces
Coverage (19 mutation commands with --dry-run):
kb.create/edit/delete agent.create/edit/delete
doc.create/upload/fetch/delete doc.delete_all (special variant)
session.delete chunk.delete profile.add/remove
auth.refresh/logout link unlink
api.{post,put,patch,delete} (api.get + --dry-run rejected, exit 2)
Envelope additions (omitempty in non-dry-run paths):
meta.dry_run: bool true when --dry-run was used
meta.plan: map {action, args} per the per-command taxonomy
Risk: metadata
--------------
cmdutil.SetRisk(cmd, action) stamps cobra.Annotations with
risk.level=destructive + risk.action=<action> on the 9 destructive
commands. The SetAgentHelp wrapper prepends a "Risk: <action>
(destructive)" line in the default text help path so agents see a clear
warning before parsing Usage. WEKNORA_AGENT_HELP=1 JSON path stays
unchanged — structured agent-help already carries warnings[].
Validation parity with the live path
------------------------------------
Every pure-local validation (flag presence, mutual exclusion, enum
bounds, URL/regex format, ResolveKBLocal for KB resolution that does not
require an SDK call) runs BEFORE the dry-run gate. This matches the
industry-standard "preview shows what live would do" contract:
--dry-run accepts exactly the same invocations the live path accepts and
rejects exactly the same invocations the live path rejects, modulo the
side-effecting work itself.
The side-effecting work (SDK calls, file writes, keyring writes, server-
side name → id resolution) is what --dry-run actually gates. Each
mutation file pairs its RunE validation block with a regression test
under *_dry_run_test.go / dryrun_validation_test.go so future refactors
don't reintroduce the gap.
Helper surface
--------------
- HandleDryRun(cmd, dryRun, plan) extracts the early-return so the
19 RunE call sites stay 3 lines each.
- EmitDryRun routes through FormatOptions.Emit, inheriting _notice /
TTY indent / --jq filtering for free.
- ResolveKBLocal mirrors ResolveKB but never calls the SDK; dry-run
paths use it so the plan reports the raw --kb value (UUID or name)
without a name → id lookup.
Streaming commands (chat, session ask, session continue-stream) are
deliberately excluded: a buffered plan makes no sense for an event
stream.
Lock semantics in the dry-run path:
- destructive + --dry-run: exit 0, no exit-10 confirmation prompt
- --dry-run + -y: byte-identical envelope to --dry-run alone
- --dry-run + --jq: filter applies to the preview envelope normally
73 lines
2.4 KiB
Go
73 lines
2.4 KiB
Go
// Package-level note:
|
|
//
|
|
// SetAgentHelp wires structured agent-targeted help onto a cobra command.
|
|
// Current coverage: chat, kb list, and destructive commands. Adding it to
|
|
// another command requires touching only that command's NewCmd (a 5-line
|
|
// copy of the existing call sites).
|
|
package cmdutil
|
|
|
|
import (
|
|
"encoding/json"
|
|
"fmt"
|
|
"io"
|
|
"os"
|
|
|
|
"github.com/spf13/cobra"
|
|
)
|
|
|
|
// AgentHelp is the structured help blob emitted when an agent invokes
|
|
// `weknora <command> --help` with WEKNORA_AGENT_HELP=1. Distinct from
|
|
// cobra's human help text — agent-readable JSON keyed by stable fields
|
|
// so an LLM doesn't need to scrape the human help table.
|
|
//
|
|
// Warnings surface both as a JSON field (agent introspection) and as an
|
|
// "AI agents:" block appended to human help. The two channels share a
|
|
// single source list so they cannot drift.
|
|
type AgentHelp struct {
|
|
UsedFor string `json:"used_for"`
|
|
RequiredFlags []string `json:"required_flags,omitempty"`
|
|
Examples []string `json:"examples,omitempty"`
|
|
Output string `json:"output,omitempty"`
|
|
Warnings []string `json:"warnings,omitempty"`
|
|
}
|
|
|
|
// SetAgentHelp attaches agent-targeted help metadata to a command.
|
|
//
|
|
// Routing:
|
|
// - WEKNORA_AGENT_HELP=1: emit the AgentHelp JSON blob to stdout and
|
|
// return — agents get clean parseable JSON, no trailing prose.
|
|
// - Otherwise (human help path): if cmd has risk annotations from SetRisk,
|
|
// prepend "Risk: <action> (<level>)" line; then render the normal human
|
|
// help via origHelp; then append the "AI agents:" Warnings block.
|
|
func SetAgentHelp(cmd *cobra.Command, ah AgentHelp) {
|
|
origHelp := cmd.HelpFunc()
|
|
cmd.SetHelpFunc(func(c *cobra.Command, args []string) {
|
|
if os.Getenv("WEKNORA_AGENT_HELP") == "1" {
|
|
emitAgentHelp(c.OutOrStdout(), ah)
|
|
return
|
|
}
|
|
// Prepend Risk: line in the default (human) branch only; the
|
|
// JSON branch above already carries warnings[]. Skip silently if
|
|
// the command has no risk annotations.
|
|
if level, action, ok := GetRisk(c); ok {
|
|
fmt.Fprintf(c.OutOrStdout(), "Risk: %s (%s)\n\n", action, level)
|
|
}
|
|
origHelp(c, args)
|
|
if len(ah.Warnings) > 0 {
|
|
w := c.OutOrStdout()
|
|
fmt.Fprintln(w)
|
|
fmt.Fprintln(w, "AI agents:")
|
|
for _, msg := range ah.Warnings {
|
|
fmt.Fprintln(w, "- "+msg)
|
|
}
|
|
}
|
|
})
|
|
}
|
|
|
|
func emitAgentHelp(w io.Writer, ah AgentHelp) {
|
|
enc := json.NewEncoder(w)
|
|
enc.SetEscapeHTML(false)
|
|
enc.SetIndent("", " ")
|
|
_ = enc.Encode(ah)
|
|
}
|