hush/vendor/github.com/prometheus/procfs/net_protocols.go
jx12n 4d9a26498e hush: one-time secret links the server cannot read
Paste a secret, get a link, send it. The first person to open it and press
Reveal sees the secret; the link dies at that moment. The recipient needs a
browser and nothing else — no account, no client, no installed tooling.

The server cannot read what it stores. AES-256-GCM happens in the browser and
the key lives in the URL fragment, which browsers never transmit, so hushd
holds ciphertext and no key material. That is a property of where the key sits
rather than a promise about our conduct, which is why there is deliberately no
endpoint accepting a plaintext secret and no server-side-encryption fallback:
two guarantees behind one URL would be worse than one honest guarantee.

Three decisions carry the design:

  * GET /s/{id} touches NO storage, not even to check existence. Slack, Teams,
    WhatsApp, iMessage and Outlook Safe Links all fetch a URL before a human
    sees it, so destroying on GET would destroy most secrets in transit and the
    recipient's "already used" would be indistinguishable from interception.
    Only POST /reveal consumes. Bot user-agent detection is an arms race;
    removing the side effect from GET is not. Pinned by
    TestGettingTheRevealPageNeverConsumesTheSecret.
  * Destruction is one Redis GETDEL, which is atomic. GET-then-DEL has a window
    where two simultaneous readers both win, and for a one-time secret that
    window is the product. The store contract demands atomicity and the same
    concurrency test runs against both implementations.
  * Missing, already-revealed, expired and evicted are ONE indistinguishable
    410. Separating them would confirm to a prober that a given link was real.

The secret id IS the capability, so secret.ID is a struct whose every
accidental path — %v, %s, String(), slog, json.Marshal — emits a redacted
handle or refuses, and the raw value needs an explicit Value(). The first
version tried to prevent leaks by implementing no String() at all; its own test
caught that Go's fmt prints unexported fields anyway, so forbidding the method
had removed the control rather than the leak.

Operationally: structured JSON on stdout in the fleet's wire format, which
Vector already collects with no annotation; six hush_* metrics on the chassis
registry with no id, IP or path in any label; five alert rules wired into
vmalert. The public Ingress enumerates /, /s/ and /api/ so /metrics, /healthz
and /readyz share the port but are unreachable from the internet — no
basic-auth middleware to maintain and get wrong.

Dependencies are vendored because go-chassis is private: the Woodpecker test
step and the in-cluster Kaniko build both run -mod=vendor with GOPROXY=off and
hold no git credential.

cmd/hush-mcp is a stdio MCP server doing the same client-side crypto locally,
so using hush from an agent preserves the same guarantee as using it from a
browser.
2026-09-03 00:08:38 -06:00

184 lines
5.3 KiB
Go

// Copyright The Prometheus Authors
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package procfs
import (
"bufio"
"bytes"
"fmt"
"strconv"
"strings"
"github.com/prometheus/procfs/internal/util"
)
// NetProtocolStats stores the contents from /proc/net/protocols.
type NetProtocolStats map[string]NetProtocolStatLine
// NetProtocolStatLine contains a single line parsed from /proc/net/protocols. We
// only care about the first six columns as the rest are not likely to change
// and only serve to provide a set of capabilities for each protocol.
type NetProtocolStatLine struct {
Name string // 0 The name of the protocol
Size uint64 // 1 The size, in bytes, of a given protocol structure. e.g. sizeof(struct tcp_sock) or sizeof(struct unix_sock)
Sockets int64 // 2 Number of sockets in use by this protocol
Memory int64 // 3 Number of 4KB pages allocated by all sockets of this protocol
Pressure int // 4 This is either yes, no, or NI (not implemented). For the sake of simplicity we treat NI as not experiencing memory pressure.
MaxHeader uint64 // 5 Protocol specific max header size
Slab bool // 6 Indicates whether or not memory is allocated from the SLAB
ModuleName string // 7 The name of the module that implemented this protocol or "kernel" if not from a module
Capabilities NetProtocolCapabilities
}
// NetProtocolCapabilities contains a list of capabilities for each protocol.
type NetProtocolCapabilities struct {
Close bool // 8
Connect bool // 9
Disconnect bool // 10
Accept bool // 11
IoCtl bool // 12
Init bool // 13
Destroy bool // 14
Shutdown bool // 15
SetSockOpt bool // 16
GetSockOpt bool // 17
SendMsg bool // 18
RecvMsg bool // 19
SendPage bool // 20
Bind bool // 21
BacklogRcv bool // 22
Hash bool // 23
UnHash bool // 24
GetPort bool // 25
EnterMemoryPressure bool // 26
}
// NetProtocols reads stats from /proc/net/protocols and returns a map of
// PortocolStatLine entries. As of this writing no official Linux Documentation
// exists, however the source is fairly self-explanatory and the format seems
// stable since its introduction in 2.6.12-rc2
// Linux 2.6.12-rc2 - https://elixir.bootlin.com/linux/v2.6.12-rc2/source/net/core/sock.c#L1452
// Linux 5.10 - https://elixir.bootlin.com/linux/v5.10.4/source/net/core/sock.c#L3586
func (fs FS) NetProtocols() (NetProtocolStats, error) {
data, err := util.ReadFileNoStat(fs.proc.Path("net/protocols"))
if err != nil {
return NetProtocolStats{}, err
}
return parseNetProtocols(bufio.NewScanner(bytes.NewReader(data)))
}
func parseNetProtocols(s *bufio.Scanner) (NetProtocolStats, error) {
nps := NetProtocolStats{}
// Skip the header line
s.Scan()
for s.Scan() {
line, err := nps.parseLine(s.Text())
if err != nil {
return NetProtocolStats{}, err
}
nps[line.Name] = *line
}
return nps, nil
}
func (ps NetProtocolStats) parseLine(rawLine string) (*NetProtocolStatLine, error) {
line := &NetProtocolStatLine{Capabilities: NetProtocolCapabilities{}}
var err error
const enabled = "yes"
const disabled = "no"
fields := strings.Fields(rawLine)
line.Name = fields[0]
line.Size, err = strconv.ParseUint(fields[1], 10, 64)
if err != nil {
return nil, err
}
line.Sockets, err = strconv.ParseInt(fields[2], 10, 64)
if err != nil {
return nil, err
}
line.Memory, err = strconv.ParseInt(fields[3], 10, 64)
if err != nil {
return nil, err
}
switch fields[4] {
case enabled:
line.Pressure = 1
case disabled:
line.Pressure = 0
default:
line.Pressure = -1
}
line.MaxHeader, err = strconv.ParseUint(fields[5], 10, 64)
if err != nil {
return nil, err
}
switch fields[6] {
case enabled:
line.Slab = true
case disabled:
line.Slab = false
default:
return nil, fmt.Errorf("%w: capability for protocol: %s", ErrFileParse, line.Name)
}
line.ModuleName = fields[7]
err = line.Capabilities.parseCapabilities(fields[8:])
if err != nil {
return nil, err
}
return line, nil
}
func (pc *NetProtocolCapabilities) parseCapabilities(capabilities []string) error {
// The capabilities are all bools so we can loop over to map them
capabilityFields := [...]*bool{
&pc.Close,
&pc.Connect,
&pc.Disconnect,
&pc.Accept,
&pc.IoCtl,
&pc.Init,
&pc.Destroy,
&pc.Shutdown,
&pc.SetSockOpt,
&pc.GetSockOpt,
&pc.SendMsg,
&pc.RecvMsg,
&pc.SendPage,
&pc.Bind,
&pc.BacklogRcv,
&pc.Hash,
&pc.UnHash,
&pc.GetPort,
&pc.EnterMemoryPressure,
}
for i := range capabilities {
switch capabilities[i] {
case "y":
*capabilityFields[i] = true
case "n":
*capabilityFields[i] = false
default:
return fmt.Errorf("%w: capability block for protocol: position %d", ErrFileParse, i)
}
}
return nil
}